This is the accessible text file for GAO report number GAO-04-171 entitled 'Microenterprise Development: USAID's Program Has Met Some Goals; Annual Reporting Has Limitations' which was released on November 17, 2003. This text file was formatted by the U.S. General Accounting Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. Report to the Chairman, Committee on International Relations, House of Representatives: United States General Accounting Office: GAO: November 2003: Microenterprise Development: USAID's Program Has Met Some Goals; Annual Reporting Has Limitations: GAO-04-171: GAO Highlights: Highlights of GAO-04-171, a report to the Chairman, House Committee on International Relations Why GAO Did This Study: Microenterprises—small businesses owned and operated by poor entrepreneurs—have potential to help the world’s poorer populations. For this reason, the U.S. Agency for International Development (USAID) included microenterprise development in its programming. In 2001, the agency reported that its was conducting microenterprise projects in 52 countries and had obligated almost $2 billion since 1988 to support its program. The program supports micro loans, among other services, to assist poor entrepreneurs. Since 1996, USAID has annually reported the program’s results. To help Congress oversee USAID’s management of its microenterprise development program, GAO was asked to (1) determine the extent to which the agency’s microfinance activities are meeting the program’s key objectives, (2) assess the reliability of USAID’s reporting on its overall microenterprise activities, and (3) examine the agency’s role in identifying and disseminating microenterprise best practices. What GAO Found: USAID’s microfinance activities have met some, but not all, of the agency’s microenterprise program objectives. These objectives are to (1) reduce poverty among participants; (2) target the poor and very poor; (3) encourage women’s participation; and (4) develop sustainable microfinance institutions (MFI). First, regarding reducing poverty— defined as alleviating its impacts or lifting and keeping a large number of people above the poverty line—GAO found that microfinance can help alleviate some impacts of poverty, incrementally improving borrowers’ income levels and quality of life and offering an important coping mechanism to poor workers and their families. However, there is little evidence that it can lift and keep many over the poverty line. Second, microfinance generally has served the poor clustered around the poverty line but not the very poor. Third, USAID has successfully encouraged the participation of women, who have comprised about two- thirds of micro loan clients since 1997. Fourth, USAID has emphasized the importance of MFI sustainability. In fiscal 2001, of 294 USAID- supported MFIs that reported on sustainability, 38 percent reported achieving full sustainability—a percentage consistent since 1999. The basic data in USAID’s Microenterprise Results Reporting (MRR) system are reliable, but certain methodological problems may affect the accuracy of some of the agency’s reporting on key program objectives. Specifically, USAID may not be reporting accurately (1) the amounts it has obligated to microenterprise activities; (2) whether 50 percent of its resources went to the very poor, as required by Congress; and (3) the sustainability of USAID-supported institutions. Further, although the agency reports annually on the activities of institutions it supports, it does not show the percentage of those institutions’ total funding that its contribution represents. USAID has identified and disseminated microenterprise best practices, providing information to its missions and implementing partners through policy guidance, training, and technical assistance. In addition, USAID has collaborated with microenterprise development provider networks and others to publish information about these practices. What GAO Recommends: GAO recommends that the Administrator of USAID review the agency’s MRR system to ensure that (1) it measures the program’s performance and (2) all data are collected, analyzed, and reported USAID completely and accurately. USAID concurred with our recommendation. [End of section] Contents: Letter: Results in Brief: Background: USAID Has Met Some Program Objectives: MRR Data Generally Reliable but May Not Accurately Reflect USAID's Microenterprise Activities: USAID Has Identified and Disseminated Best Practices: Conclusion: Recommendation to USAID: Agency Comments and Our Evaluation: Appendix I: Scope and Methodology: Appendix II: Microenterprise Best Practices: Appendix III: Review of Microenterprise Studies: Appendix IV: USAID's Microenterprise Results Reporting: Appendix V: Comments from the U.S. Agency for International Development: GAO Comments: Appendix VI: GAO Contact and Staff Acknowledgements: GAO Contact: Staff Acknowledgements: Tables: Table 1: Findings and Recommendations from 22 Selected Studies on Microfinance and Microenterprise Development: Figures: Figure 1: USAID's Microenterprise Development Program, Highlighting Microfinance Activities: Figure 2: Conceptual Diagram of Populations along the Poverty Scale: Figure 3: Microentrepreneurs in Bulgaria: Figure 4: Microentrepreneurs with Loans from Various USAID-funded MFIs: Figure 5: Financial Sustainability of Institutions, by Average Loan Size: Figure 6: Clients in USAID-funded Micro Loan Projects, by Gender, 1997- 2001: Figure 7: Women Entrepreneurs at USAID-funded MFIs in Peru: Figure 8: Financial Sustainability of MFIs, by Average Number of Borrowers: Figure 9: Available USAID Obligations Data for Poverty Lending, Fiscal Years 2000 and 2001: Abbreviations: ACCION: Americans for Community Cooperation in Other Nations: AIMS: Assessing the Impact of Microenterprise Services: BDS: Business Development Services: CGAP: Consultative Group to Assist the Poor: CRS: Catholic Relief Services: GAO: U.S. General Accounting Office: GEMINI: Growth and Equity through Microenterprise Investments and Institutions: MFI: microfinance institution: MRR: Microenterprise Results Reporting: NGO: nongovernmental organization: SEEP: Small Enterprise Education and Promotion Network: USAID: U.S. Agency for International Development: United States General Accounting Office: Washington, DC 20548: November 17, 2003: The Honorable Henry Hyde: Chairman: Committee on International Relations: House of Representatives: Dear Mr. Chairman: Microenterprises--small businesses owned and operated by poor entrepreneurs--have potential to help the least advantaged populations in the developing world. Recognizing this potential, the U.S. Agency for International Development (USAID) has included microenterprise development in the agency's strategy for economic development and poverty reduction. In 2001, the agency reported that it was conducting microenterprise projects in 52 countries and had obligated almost $2 billion since 1988 to support its program. The Microenterprise for Self-Reliance and International Anti-Corruption Act of 2000,[Footnote 1] enacted on October 17, 2000, expressed strong congressional support for microenterprise projects, stated policy preferences for implementing such projects, and authorized increased funding. The key objectives of USAID's microenterprise development program are to reduce poverty among microenterprise owners, workers, and families;[Footnote 2] target the poor and the very poor;[Footnote 3] encourage participation by women; and develop sustainable microfinance institutions (MFI).[Footnote 4] The majority of USAID's microenterprise funding supports microfinance services, particularly micro loans. Since 1996, USAID has annually surveyed missions and implementing institutions regarding their microenterprise activities and issued a report to Congress. To assist your committee in overseeing the agency's management of its microenterprise development program, you asked that we (1) determine the extent to which USAID's microfinance activities are meeting the program's key objectives, (2) assess the reliability of the agency's reporting on its overall microenterprise activities, and (3) examine the agency's role in identifying and disseminating microenterprise best practices. To identify key program objectives, we reviewed USAID's policy guidance and the Microenterprise for Self-Reliance and International Anti- Corruption Act of 2000. To determine whether USAID is meeting these objectives, we interviewed officials from USAID, USAID's implementing partners, and experts; conducted fieldwork in Peru, Egypt, and Bulgaria, three countries with established USAID microenterprise programs; and reviewed academic and applied studies of microenterprise projects and outcomes. Because most available USAID data and research literature are focused on microfinance, particularly on micro loans, we primarily analyzed this aspect of the agency's microenterprise program. To assess USAID's microenterprise reporting, we evaluated the efforts of the contractor responsible for the reporting, analyzed the electronic files containing program data, and checked the reliability of data in the contractor's system and in the countries we visited. To determine USAID's role in identifying and disseminating information on best practices, we reviewed a wide body of literature on the subject; interviewed USAID officials, implementing partners, and a panel of experts; and reviewed relevant documents in Washington, D.C., and the countries where we conducted fieldwork. (See app. I for a discussion of our methodology.): Results in Brief: USAID's microfinance activities have met some, but not all, of the program's key objectives. First, regarding the reduction of poverty, microfinance can help alleviate some impacts of poverty, incrementally improving borrowers' income levels and quality of life and providing poor individuals, workers, and their families with an important coping mechanism. However, there is little evidence in the research literature or consensus among experts that microfinance alone can lift and keep large numbers of people above the poverty line. Second, although USAID- funded studies and academic analyses have found that microfinance activities generally serve the poor clustered around the poverty line, few loans appear to be reaching the very poor. At the same time, some research suggests that loans to the very poor may be ineffective, creating unmanageable debt, unless combined with other services. Third, the program has encouraged women's participation in microfinance projects: Women have comprised two-thirds or more of the micro loan clients in USAID-funded microfinance projects since 1997, according to USAID data. Finally, the agency has emphasized the importance of developing fully sustainable MFIs, and some of the MFIs it funds report having made progress toward this goal. Most of the program data in USAID's Microenterprise Results Reporting (MRR) system were generally reliable, but certain methodological problems may prevent the agency from accurately reporting its progress in meeting key objectives. Because of the way the MRR obligations data are collected, USAID cannot report microenterprise funding levels with certainty. In addition, its method of estimating the percentage of loans to the very poor prevents the agency from stating with certainty that it has met the requirement that 50 percent of its microenterprise resources go to the very poor. Finally, because differing definitions are used to calculate sustainability, the sustainability data that USAID-supported MFIs provide may not be reliable. Moreover, although the agency's annual reports describe the overall activities of institutions that receive USAID support, the reports do not provide sufficient data on USAID's contribution to MFIs and other service providers. USAID has identified and disseminated best practices for microenterprise development. The agency has funded initiatives to identify best practices and has provided information on these practices to its missions and implementing partners through policy guidance, training, and technical assistance. USAID also has collaborated with networks of microenterprise implementing partners and other organizations to publish information on best practices. (App. II contains a list of key best practices.): We are recommending that the Administrator of USAID review the agency's MRR system with the goal of ensuring that its annual reporting is complete and accurate. Specifically, the Administrator should review and reconsider the methodologies used for the collection, analysis, and reporting of data on annual spending targets, outreach to the very poor, MFI sustainability, and the contribution of USAID's funding to the institutions it supports. USAID provided written comments on a draft of this report (see app. V for the agency's comments and our detailed response). USAID concurred with our recommendation that it make improvements in its microenterprise reporting, but it took issue with three points: (1) USAID stated that its microfinance efforts have reached the very poor, as statutorily defined, based on the number of micro loans made by institutions that it supports. This report recognizes that loan size is considered by USAID and the microenterprise industry as an inadequate indicator of whether micro loans are reaching the very poor, and Congress has directed the agency to develop more precise poverty measurement tools. Our report reflects current research and consensus on the extent to which the very poor are being served by the program; (2) USAID also stated that it uses a single definition of sustainability and that the sustainability data reported in the MRR are reliable. However, we documented differing definitions and interpretations that affect the reliability of reported sustainability data; and (3) USAID stated that it would be difficult to allocate the microenterprise accomplishments reported in the MRR between USAID and other donors, but it agreed to make other improvements to the MRR, including providing more explicit instructions on activities to include in reporting and taking other steps to improve data accuracy. If USAID implements these improvements, which we believe are responsive to our recommendation, it could help improve the reliability of the information in the MRR. Background: Microenterprise development and its primary component, microfinance, emerged in the 1980s. Microfinance--the supply of micro loans,[Footnote 5] savings, and other financial services to the poor--has operated on the premise that the poor will invest loans in microenterprises, repaying the loans out of profits, and the enterprises will grow, potentially lifting large numbers of people out of poverty. Microfinance practitioners have demonstrated that the poor will repay loans on time and that, despite high transaction costs, micro loans can support financially viable lending institutions. Through the 1990s, a wide variety of institutions operating in many countries have used this model, adapting it to local conditions. By USAID's definition, a microenterprise consists of a poor owner- operator and nine or fewer workers. Microentrepreneurs--typically small shopkeepers, craftspeople, and vendors--face a range of impediments to improving their standard of living. Most rarely have sufficient collateral to meet loan requirements at traditional banks; according to one report, only 2.5 percent of potential microenterprise operators have access to financial services other than moneylenders.[Footnote 6] USAID's microenterprise development program focuses on three key areas: * helping establish and fund MFIs to provide loans and other financial services to the poor and very poor, * funding business development services (BDS) to help improve the business skills of microentrepreneurs and develop markets for microenterprises, and: * advocating for host government policy reforms to enhance microenterprise development. USAID has been the leading bilateral donor of funds and technical assistance to microenterprise development projects since 1988, when it began formally tracking funding. USAID provides funding through nongovernmental organizations, contractors, and, occasionally, government organizations that implement the client-level activities. USAID reported that it committed $158.7 million to microenterprise development activities in fiscal year 2001, compared with $164.3 million in fiscal year 2000. Almost two-thirds of USAID's microenterprise development obligations in fiscal year 2001 were directed to providing financial services, principally for creating and strengthening existing credit and savings institutions; the other third went to BDS and policy activities. Figure 1 shows an overview of USAID's microenterprise program, highlighting its microfinance component. In fiscal year 2001, USAID-assisted institutions served more than 2.8 million loan clients and more than 3.5 million savings clients. USAID-assisted institutions also provided BDS to more than 800,000 microenterprise clients, including market research, new product development and testing, technology development, business counseling, and marketing assistance. Figure 1: USAID's Microenterprise Development Program, Highlighting Microfinance Activities: [See PDF for image] [End of figure] USAID's microenterprise development program, which is highly decentralized, funded projects in 52 countries in 2001. USAID's missions design, implement, and monitor the microenterprise projects, obligating about 80 percent of the program's funding. In Washington, D.C., the Microenterprise Development Division provides policy guidance and manages a number of grants and, along with other USAID offices, provides about 20 percent of the agency's microenterprise funding. The division also provides technical support for the missions and conducts research on microenterprise issues. USAID takes a collaborative approach in its microenterprise development program. For example, USAID policy is to identify, support, and strengthen existing MFIs with established performance records to help meet its microenterprise objectives. It also funds studies of the MFIs it has supported, to assess impacts and to identify best practices for both USAID and the entire microenterprise industry. One major USAID research project, Assessing the Impact of Microenterprise Services (AIMS), was initiated in 1995. Most of the AIMS studies focused on a single country or activity. However, two recent AIMS studies examined USAID-and non-USAID-funded microenterprise activities in seven countries[Footnote 7] and synthesized key findings from a number of other studies,[Footnote 8] respectively. USAID's microenterprise development program has targeted the poor and the very poor (see fig. 2). The poor are those whose annual income is at or below the poverty line as defined by the host country. The very poor are those with an annual income 50 percent or more below the poverty line as established by the government of the country. In addition, the vulnerable nonpoor, who also receive microenterprise assistance, are those whose annual income is just above the poverty line. Figure 2: Conceptual Diagram of Populations along the Poverty Scale: [See PDF for image] Note: This diagram is for illustrative purposes only and is not drawn to scale. The national poverty line is defined by each country. [End of figure] Since 1994, USAID's policy has been to devote half of its microenterprise development resources to activities targeting the very poor. In 2000, this policy was established as law by the Microenterprise for Self-Reliance and International Anti-Corruption Act of 2000, which required that 50 percent of all microenterprise resources be targeted to very poor entrepreneurs. From 1994 until the implementation of the 2000 legislation, USAID defined loans to the very poor, or "poverty loans," as those with an average balance of $300 or less per borrower (in 1994 dollars). The 2000 legislation established the loan level at $1,000 or less for Europe and Eurasia, $400 or less for Latin America, and $300 or less in the rest of the world. USAID annually collects data on its microenterprise program through the MRR. (See app. IV for a discussion of the MRR methodology.) Data are collected through surveys of USAID staff in headquarters, overseas missions, and institutions that receive USAID funding. USAID staff provide the data on funding obligations, and implementing institutions provide programmatic data on number and gender of clients, number of loans, amount of lending to the very poor, MFI sustainability self- assessments, BDS services provided, and policy initiatives. In addition to the USAID-supported studies, a large body of published and unpublished work has been generated by what is commonly referred to as the microfinance industry. This research includes empirical studies, a broad array of theoretical analyses, and materials on best practices. However, because the industry is relatively new, large gaps exist in important research areas. For example, experts generally agree that reliable and valid impact assessments are lacking or limited in scope. (See app. III for a review of 22 microenterprise studies.): USAID Has Met Some Program Objectives: USAID's microfinance activities have achieved some of the agency's key program objectives but made limited progress toward others. First, microfinance can help to alleviate the effects of poverty among participants. However, there is less evidence that microfinance has lifted or kept large numbers above the poverty line as established by host countries. Second, USAID's program has generally reached the poor but not the very poor. Third, the program appears to have succeeded in encouraging the participation of women through micro loans. Finally, USAID has emphasized the importance of developing sustainable MFIs and has made some progress toward this goal. Microfinance Can Help Alleviate Some of Poverty's Impacts: Our fieldwork and review of USAID-funded studies and research literature showed that microfinance can alleviate some of the impacts of poverty among recipients.[Footnote 9] USAID officials, implementing partners, and borrowers also concurred that it can help to mitigate some effects of poverty. Most borrowers said that the microfinance institutions were their only formal source of credit or savings and stated that the program helped improve their lives, incomes, businesses, and sense of self-worth. (See fig. 3.) The broader academic studies of microfinance we reviewed, as well as the experts we consulted, generally agreed that micro loans can help poor individuals, workers, and their families cope with personal and economic shocks, such as illness or death of family members, and manage risks associated with living in poverty. In addition, USAID officials noted that microfinance projects are demand driven, put money directly into the hands of the borrowers, and generally have loan repayment rates close to 100 percent. The experts with whom we spoke agreed that microfinance programs can have other benefits, such as helping to build a nation's financial sector. Figure 3: Microentrepreneurs in Bulgaria: [See PDF for image] [End of figure] The studies we reviewed and experts we consulted also posited that microfinance can help increase microentrepreneurs' working capital, thereby enhancing their household income. However, positive impacts of specific microfinance projects have been limited in scope and have varied according to economic, social, and market conditions as well as the design and aims underlying particular programs. USAID-funded impact studies in two countries we visited, Peru and Egypt, found some positive effects from microfinance. For example, the study in Peru found that the program helped increase assets at the microenterprise level; it also noted some positive effects at the household and individual levels. According to the study, economic recession in Peru during 1997 and 1999 may have limited the effects of the small loans.[Footnote 10] A study in Egypt found that a microfinance program offering group lending to poor women helped its clients establish and expand their businesses.[Footnote 11] The loans also enabled them to improve their standard of living by contributing to the household budget, renovate their homes, and provide their children with a better education through private tutors. (See fig. 4.): Figure 4: Microentrepreneurs with Loans from Various USAID-funded MFIs: [See PDF for image] [End of figure] Although microfinance can help alleviate some of the impacts of poverty, the research literature does not show conclusively that it has lifted large numbers of people above the poverty line. In our review, we identified two studies on this issue. The 2002 AIMS study--examining microfinance projects in three countries, two of which were USAID supported--found that "there was no dramatic movement of client households out of poverty over the two-year span of the study."[Footnote 12] The second study analyzed the three major microfinance programs in Bangladesh--two of which received USAID funding in fiscal year 1998--and found that about 5 percent of program participants per year rose above the poverty line during the period covered by the study.[Footnote 13] The experts we consulted generally agreed that although microfinance can help to alleviate some of poverty's impacts, too few long-term studies have been conducted to determine whether microfinance can lift, and keep, significant numbers of clients above the poverty line. These experts also emphasized that because the poverty line is a problematic and somewhat artificial measure, most impact studies have not focused on estimating the number of borrowers who cross and remain above it. The experts and practitioners we interviewed and whose work we reviewed now generally conclude that microfinance alone is not sufficient to lift large numbers of people out of poverty. The challenges the poor face--limited education, few opportunities, legal and cultural barriers--are difficult to overcome with micro loans. Moving out of poverty usually requires a combination of strategies by different household members, and, according to a USAID program official, "backsliding is possible and even frequent.": Most USAID Funding Has Reached the Poor but Not the Very Poor: Although the agency's microfinance activities serve the poor, they generally appear not to reach the very poor, according to our review of USAID studies and the research literature. In addition, as mandated by the Microenterprise for Self-Reliance and International Anti- Corruption Act of 2000, the agency uses small loan size as an indicator of loans to the very poor; however, this is now generally considered an inadequate measure of success in reaching that population. Moreover, some evidence suggests that micro loans can have unintended negative consequences among very poor borrowers. Finally, meeting the requirement of targeting 50 percent of microenterprise resources to the very poor could hamper MFI sustainability. USAID's Microfinance Activities Generally Have Not Reached the Very Poor: USAID studies and other research literature on microfinance show that microfinance activities serve those clustered just above and below the poverty line but generally do not reach the very poor. According to the 2002 USAID-funded AIMS study, based on work in three countries, both the vulnerable nonpoor and the poor participate in the program, with the very poor making limited use of USAID-supported microfinance services. The 2000 AIMS study[Footnote 14] reported that in the projects studied in four countries, the majority of clients were poor, followed by the vulnerable nonpoor. This study also found that approximately 40 percent of USAID microfinance clients in Bangladesh were very poor but that in Bolivia, the Philippines, and Uganda, the number of very poor ranged from "almost none" to "some," although it did not quantify the precise numbers. In addition, the 2000 AIMS study noted that 20 other microfinance impact studies had found limited participation by the very poor. The broader literature on microfinance confirms that the microfinance industry has reached the poor and vulnerable nonpoor but relatively few of the very poor. For example, one widely cited study[Footnote 15] found that microfinance lenders in Bolivia tended to serve those near the poverty line, not the very poor. During our fieldwork, representatives from USAID and their implementing partners told us, based on their experience with the program, that few loans went to the very poor--a finding generally consistent with academic studies of projects in other countries. USAID officials in the countries we visited said that the very poor rarely take out loans because they may lack the economic opportunities to repay the loans and are reluctant to increase their debt levels. According to the 2000 AIMS study, not enough information is available to determine whether (1) the very poor choose not to borrow to avoid additional debt; (2) MFI staff disqualify the very poor because of concern over their ability to repay the loans; or (3) other types of loans and services, such as savings or insurance, would better meet the needs of the very poor. Small Loan Size an Inadequate Measure of Whether Loans Reach the Very Poor: Although most MFIs use small loan size as an indicator of loans to the very poor (as mandated in the 2000 act), in practice this is an inadequate method. It is based on the assumption that small loans appeal only to the very poor, and it is widely used in part because it is easy to administer. However, many practitioners, including USAID, now generally consider loan size an inadequate indicator of clients' level of poverty. In June 2003, legislation was enacted amending the Microenterprise for Self-Reliance and International Anti-Corruption Act of 2000 to ensure the development of more precise poverty measurement tools.[Footnote 16] The amendments required USAID to develop, test, and certify at least two low-cost methods for determining recipients' poverty level by October 1, 2004, and begin using one of the methods by October 1, 2005. The amendments also expanded the definition of the very poor to specifically include those living on less than $1 per day. Loans to Very Poor May Be Ineffective Unless Combined with Other Services: Although some evidence suggests that micro loans may help alleviate the impacts of poverty, evidence also suggests that in some cases these loans may affect very poor borrowers more negatively than positively and may be more effective in combination with complementary services. Within the microfinance industry, little consensus exists about the effectiveness of micro loans to the very poor. USAID officials in the countries we visited stated that economic and social impediments in those countries often make loan repayment difficult for the very poor. In Peru, a representative of a large U.S.-based implementing partner told us that her organization typically does not lend to the very poor, considering social services, not loans, more appropriate for that population. In Egypt, one of the largest USAID-supported MFIs said that it has started a separate grant program to reach the very poor. USAID officials in Bulgaria we visited said that the poor were more able than the very poor to expand their enterprises and, as a result, to hire the very poor. In addition, a USAID program official stated that microfinance might not always be an appropriate intervention for the very poor, since they often cannot use the loans productively. Some research also indicates that micro loans alone may not be an appropriate assistance mechanism for people below a certain level of poverty because they may increase their debt to unmanageable levels. Other research has attempted to show that with a strong commitment to reaching the very poor, and with a well-targeted program attuned to the needs of very poor clients, microfinance can have positive impacts.[Footnote 17] At the same time, recent studies suggest that to reach the very poor, the microfinance industry needs to move beyond loans and offer the very poor other services, such as savings and insurance. [Footnote 18] For example, a 2002 strategic evaluation of the Consultative Group to Assist the Poorest (CGAP)[Footnote 19] stated that savings may be the most important financial service for the very poor, since it provides a way to accumulate money without risking debilitating indebtedness. In addition, the 2002 AIMS report and other research indicated that, because of difficulties in reaching the very poor with micro loans and the potential for indebtedness, there is a need to expand the type of products or assistance targeted to this group. These products can include savings, insurance, and money transfer services; nonfinancial business development services; and reforms of key policies, programs, institutions, and regulations that can affect the very poor. Last, a 2003 CGAP publication states that donor funding for microfinance should complement, not substitute, for investments in core services, like health, education, and infrastructure,[Footnote 20] a view that also reflects USAID's policy, according to agency officials. Poverty Lending Requirement May Hamper MFI Sustainability: USAID officials stated that implementing the requirement that 50 percent of funds be targeted to the very poor, based on the loan sizes set by the 2000 act, could make individual MFI sustainability more difficult to achieve. Officials at the missions we visited said that their primary objective was to develop sustainable MFIs. In Bulgaria, officials with USAID and its implementing partners, Catholic Relief Services (CRS) and Opportunity International, said that imposing this requirement on individual MFIs could create unsustainable institutions, because managing a high percentage of small loans would increase costs associated with servicing these loans. The mission in Egypt, which began its microfinance program in 1988, did not offer poverty lending until 1999, when it judged that the institutions it supported were financially viable and stable enough to begin making such loans.[Footnote 21] The research literature we reviewed also indicates that MFIs that are considered financially sustainable generally do not reach the very poor in large numbers. Our analysis of data in the MicroBanking Bulletin, a publication in which MFIs report financial and programmatic information, indicate a direct correlation between larger average loan size and increased financial sustainability (see fig. 5). Further, according to a CGAP assessment, donor confidence during the mid-to-late 1990s that most MFIs could both reach the very poor and become sustainable has since declined. Figure 5: Financial Sustainability of Institutions, by Average Loan Size: [See PDF for image] Note: MFIs are considered financially sustainable when they can cover 100 percent of their costs, as well as the cost of borrowing funds at commercial rates. [End of figure] A USAID program official said that the poverty lending requirement can work against the goal of developing sustainable MFIs, since it directs the agency to target half of its resources to those may be least able to repay the loans. The official added that by focusing its resources on the poor and the vulnerable nonpoor, who can use loans more productively, the agency could increase the likelihood of developing sustainable institutions. USAID's Microfinance Activities Have Reached Many Women: USAID data indicate that the agency succeeded in reaching large numbers of women clients through its microcredit activities in fiscal years 1997 to 2001 (see fig. 6).[Footnote 22] Figure 6: Clients in USAID-funded Micro Loan Projects, by Gender, 1997- 2001: [See PDF for image] [End of figure] The broader research literature we reviewed shows that micro credit activities have successfully targeted women. Generally, the literature suggests that female clients have had better loan repayment rates and lower default rates than male clients. Microcredit services are of considerable importance to poorer women, who tend to have more limited access to other financial services than men. Research also shows that micro loans have generally improved female clients' participation in decision making at the household and business levels.[Footnote 23] Our fieldwork indicated that USAID-supported MFIs' focus on women varied by country and project type. In fiscal year 2001, more than two- thirds of the USAID-funded MFI micro loan clients in Peru were women, and in Egypt and Bulgaria, just under half of those clients were women. Within these countries, we found that project design affected women's participation rates. Projects employing group lending[Footnote 24] or offering nonfinancial incentives, such as health care, tended to have a higher percentage of female clients. For example, as a result of group lending projects that began in 1999 for women from poor communities, the overall percentage of women clients across USAID-funded microfinance activities in Egypt increased from 17 percent in fiscal year 2000 to about 45 percent in fiscal year 2001. In Peru, MFIs such as ProMujer, whose clients are nearly all women, offer borrowers group loans, day care, health education, and medical referral services. (See fig. 7.): Figure 7: Women Entrepreneurs at USAID-funded MFIs in Peru: [See PDF for image] [End of figure] Some Progress Made toward MFI Sustainability: USAID has emphasized developing sustainable MFIs, and available data suggest that some progress has been made toward this goal. In fiscal year 2001, of the 294 USAID-funded MFIs that reported sustainability levels, 112 stated that they had achieved full sustainability. USAID does not collect sustainability data from MFIs that no longer receive funding. As a result, it lacks the long-term data needed to determine whether MFIs it has supported have continued to provide services in a sustainable manner and if so, for how long. The research literature we reviewed indicates that achieving full sustainability has been difficult for the broader microfinance industry. Further, the literature indicates that fully sustainable MFIs tend to reach larger numbers of borrowers. Finally, MFI sustainability can be transient, subject to factors such as mismanagement and economic shocks. USAID Has Emphasized MFI Sustainability: Within the microfinance industry, USAID is a leader in promoting MFI sustainability. Before receiving USAID funding, an MFI must provide a plan outlining the major steps it will take to achieve sustainability. USAID expects MFIs to attain full sustainability within 7 years of receiving USAID assistance. USAID and other donors consider sustainability to be an important goal because it requires that MFIs manage operations efficiently and meet clients' needs consistently. Further, achieving sustainability allows institutions to continue providing services after donor funding ceases. According to one CGAP official, "Aiming for sustainability is paramount.": USAID determines an MFI's progress in achieving full sustainability by using an interim measure it calls operational sustainability. An MFI is considered operationally sustainable if revenues from interest and fees cover all of its operational expenses, including salaries and other administrative expenses. To be considered fully sustainable, the organization must cover both its operational and financial costs, such as the cost of borrowing funds at commercial interest rates, while taking into account inflation and any subsidies. USAID-Funded MFIs Report Some Progress toward Full Sustainability: Available data suggest that USAID-supported MFIs have made some progress toward achieving full sustainability. In fiscal year 2001, 294 USAID-supported MFIs reported sustainability levels;[Footnote 25] of these, 112, or 38 percent, said that they were fully sustainable, a percentage that had remained consistent since 1999.[Footnote 26] Because USAID does not monitor MFI sustainability once its funding stops, it lacks long-term data to determine whether the MFIs it has supported continue to be sustainable, and if so, for how long. To assess trends in MFI sustainability, we analyzed MRR data from 1995 to 2001; 45 MFIs reported sustainability data for both 1995 and 2001. Of these, 15 (33 percent) reported reaching full sustainability at the time of the 2001 survey. These percentages are similar to the percentage reported for a 2002 MicroBanking Bulletin survey of established MFIs[Footnote 27] but higher than those reported for the overall microfinance industry. According to officials from CGAP, the Foundation for International Community Assistance, and Americans for Community Cooperation in Other Nations (ACCION), of the approximately 10,000 MFIs currently operating, the number that are sustainable or are expected to survive in the long term ranges from an estimated "few dozen"[Footnote 28] to 250. However, those MFIs that are currently reported as sustainable serve about 80 percent of the total microfinance clients worldwide, according to these officials. The experts we interviewed agreed that the majority of microfinance clients are served by a few large, sustainable MFIs. In some cases, USAID has continued to fund MFIs that reported achieving full sustainability and MFIs that did not achieve sustainability within 7 years of receiving USAID funding. USAID officials said that the primary reason for continuing to fund these MFIs is to expand microfinance services to new areas. For example, in Egypt, one of the institutions listed as financially sustainable has received USAID funding for 14 years to support expansion, according to mission officials. In Bulgaria, an institution that had not attained operational sustainability received USAID funding for fiscal years 1995 to 2003 and is expected to continue receiving support until fiscal year 2006, when USAID is expected to end its microenterprise activities in that country. USAID officials in Bulgaria said that the country's macroeconomic and financial instability, along with regulatory and legal hurdles, has adversely affected MFIs. The research literature we reviewed indicates that a large majority of existing MFIs are not, and are not expected to become, fully sustainable. The literature further indicates that MFIs with a large number of clients have higher levels of financial self-sufficiency and profitability than smaller MFIs. For example, Bank Rakyat Indonesia's Micro Division had about 2.8 million borrowers and 27 million savings depositors in fiscal year 2001 and has reported full financial sustainability since the early 1990s. In addition, data reported in the MicroBanking Bulletin, based on financial and portfolio data of leading microfinance institutions worldwide, indicate that institutions with a large loan portfolio and number of clients have higher levels of financial sustainability than smaller institutions (see fig. 8). Figure 8: Financial Sustainability of MFIs, by Average Number of Borrowers: [See PDF for image] Note: MicroBanking Bulletin defines MFI size by the average number of borrowers. [End of figure] MFIs are considered financially sustainable when they can cover 100 percent of their operating costs, as well as the cost of borrowing funds at commercial rates. MFI Sustainability Can Be Transient: In analyzing the status of 81 USAID-funded MFIs that reported on financial sustainability over a 5-to 7-year period, we found that over one-fifth (18) reported achieving full sustainability but later reported that they were no longer fully sustainable. MFI sustainability can change rapidly because of various factors, as the following examples from Peru illustrate. * PRISMA, one of the largest USAID-funded MFIs, became unsustainable because of mismanagement and the theft of $2 million by employees. (USAID officials said that steps had been taken to recover the funds and that PRISMA will remain ineligible for future support until this situation is resolved.): * CARITAS, an MFI affiliated with Catholic Relief Services, experienced declines in full sustainability at five[Footnote 29] of its eight branches during a 19-month period. Over the next 4 months, the sustainability of two branches improved, but during the subsequent 23- month period, lenient loan repayment practices at three branches resulted in a significant decline in sustainability and a consequent decline in portfolio quality at those branches. External factors such as hyperinflation, host country policy, and regulatory environment can also significantly affect MFI sustainability. For example, a 2002 USAID-funded study concluded that in Bulgaria, the effective ability of the three USAID-supported MFIs "to continue probably expires with the ending of their donor program unless their full [legal] . . . and regulatory status is resolved."[Footnote 30] The study noted that for the most promising of the three MFIs, Bulgaria's restriction on attracting savings from borrowers or others was a major barrier to sustainability. In addition, a USAID official stated that the devaluation of the U.S. dollar, among other factors, contributed to a 14 percent drop in operational sustainability at this MFI between fiscal year 2001 and fiscal year 2002, because loans were disbursed in dollars but collected in the local currency. MRR Data Generally Reliable but May Not Accurately Reflect USAID's Microenterprise Activities: Although the basic data collected for USAID's MRR are generally reliable, certain methodological problems may impede accurate reporting on the agency's progress in meeting key goals. Specifically, it may not be reporting accurately (1) the actual amounts obligated to microenterprise activities, (2) whether 50 percent of USAID's resources went to the very poor, and (3) the sustainability of USAID-supported MFIs. Moreover, although the annual MRR reports on the overall activities of MFIs that receive any USAID monies, it does not provide sufficient data on USAID's contribution to MFIs and other service providers. Basic MRR Data Are Generally Reliable: We assessed the reliability of basic MRR data in terms of accuracy, completeness, and consistency and found that they generally met these criteria (see app. I for our methodology and app. IV for a discussion of the MRR). These data include the number of clients, the percentage of women clients, and the dollar amounts of the institutions' portfolios. USAID collects most of the data via surveys filled out by the institutions receiving USAID assistance. According to the contractor responsible for collecting and analyzing the MRR data, the survey questions for institutions were pretested and should be understood by respondents. (The survey is available in English, Spanish, and French.): The contractor and USAID officials stated that they review the data for completeness, accuracy, and consistency. MRR staff reported that they compared current and past year survey responses to identify inconsistent responses and investigated these responses as warranted. Although two of the three USAID missions we visited did not perform the checks recommended by USAID's policy guidance, most of the data we examined were generally accurate. Reliability of Obligations Data Is Uncertain: We observed several problems with the reporting of USAID's obligations that may affect the reliability of the data in the MRR. First, the 2001 MRR publication includes obligations for many clients and institutions that do not meet the MRR's definition of a microenterprise as "compris[ing] 10 or fewer employees, including unpaid family workers, in which the owner/operator of the enterprise . . . is considered poor."[Footnote 31] For example, of the roughly 120 institutions that reported BDS and policy development programs in the 2001 MRR publication, more than half reported serving some clients whose incomes were above the national poverty line or who owned businesses that were not microenterprises. In addition, at least one-third of all BDS clients included in the MRR in 2001 had estimated incomes above the poverty line. Furthermore, in Peru, an institution that had received an obligation of $1.2 million in 2001 reported all its clients to the MRR, even though only about one-third of its clients were microentrepreneurs. Finally, almost a quarter (12 of 42) of the USAID- supported MFIs in Eastern Europe reported to the MRR loans exceeding $10,000, despite the regional loan size limit of $10,000. Second, underreporting of microenterprise obligations in the MRR may occur. According to the USAID contractor, some of the missions that report may not list all obligations for microenterprise activities. For example, in 2000, the contractor who collects the MRR data found $7 million of underreported microenterprise obligations, which was subsequently included in the obligations totals. In addition, the MRR does not track expenditures for its microenterprise programs, and obligations reported to the MRR may not accurately reflect actual program expenditures. During our fieldwork, we found situations where obligations differed significantly from actual expenditures. For example, for fiscal years 1991 to 2000, the MRR reported obligations of about $160 million to microenterprise programs at the USAID mission in Egypt. However, mission officials told us that the program actually spent about 50 percent of this obligation. Estimate of Obligated Funds Going to the Very Poor May Be Inaccurate: USAID reported that 53 percent of its obligated microenterprise funds went to the very poor in fiscal years 2000 and 2001. However, our analysis indicates that the MRR may not accurately estimate the percentage of microenterprise development funding that is targeted to the very poor. We found the following limitations: * The MRR lacks information on poverty lending for a significant portion of total microenterprise obligations. In fiscal year 2000, it had data for only 32 percent of obligations, and for fiscal year 2001, it had data for only 41 percent of obligations (see fig. 9). Figure 9: Available USAID Obligations Data for Poverty Lending, Fiscal Years 2000 and 2001: [See PDF for image] Note: In fiscal year 2000, an estimated $28.5 million in loans went to the very poor, and in fiscal year 2001, an estimated $34.2 million in loans went to the very poor. [End of figure] * USAID's method for calculating overall poverty lending extrapolates from the available data and assumes that institutions that did not respond to the MFI survey provided the same amounts of poverty lending as those that did respond. However, unlike the respondents, many of the nonrespondents did not make loans or performed activities that were not directly involved with poverty lending. Nonrespondent activities, which totaled $94 million--roughly 60 percent of the total fiscal year 2001 obligations--included a range of services. For example, in 2001, USAID obligated about $5 million to support its microenterprise staff and about $2 million for research and other support activities. * Many BDS programs that report on outreach to the very poor are likely to provide inaccurate data. While MFIs report the dollar value of poverty loans they have made, many BDS providers must estimate and report the number of their clients who have received poverty loans from any source--data that, according to a USAID program official, the BDS providers often lack. According to a USAID program official, the agency went to considerable effort to collect data from institutions that make poverty loans. USAID officials acknowledged that it is difficult to estimate future poverty lending for institutions that have not yet begun to make poverty loans and those that provide services rather than loans to clients. However, USAID's annual MRR report does not inform the reader of the extent or impact of these limitations. MFI Sustainability Data May Not Be Reliable: Because USAID-supported MFIs use different definitions to calculate sustainability, the sustainability data reported in the MRR may not be reliable. USAID supplies differing definitions of sustainability, one for the MRR and one for its Implementation Grant Program awards. In addition, not all MFIs reporting to the MRR use the definition suggested there; for example, one MFI with affiliates in more than 20 countries requires its affiliates to report sustainability to the MRR using a more stringent definition. Further, MFIs can and do interpret the underlying MRR definition of sustainability differently; for example, some basic terms such as "financial costs" are not defined and are subject to various interpretations. As a result, the contractor responsible for collecting and analyzing these data stated that they should not be considered reliable. MRR Does Not Provide Sufficient Data on USAID's Contributions: The MRR does not provide information on USAID's level of contributions to MFIs and other service providers it supports, making it difficult to determine the scope of the agency's microenterprise development funding. The MRR reports an MFI's total number of clients and loans, regardless of the level of USAID's contribution to that MFI. For example, in 2000, USAID obligated $400,000 to an MFI in Ecuador that reported loans of $80 million and $477,000 to an MFI in Senegal that reported loans of $336,000. As a result, the annual report lists a large number of clients, loans, and other activities that were not funded by USAID and in many cases were funded by other donors, foundations, and private individuals. In addition, USAID requires institutions that provide technical assistance to MFIs to complete the MFI survey. Because these technical assistance providers make no loans, reporting the number of loans and clients served by the MFIs they assist may provide an inaccurate impression of USAID's micro lending activities. For example, we found that the institution listed as in the MRR as the largest lender in Peru in 2001 did not make any loans or serve any clients directly. Instead, it provided technical assistance to more than 20 MFIs. However, the MRR reported that this institution had a $37 million loan portfolio and 20,000 clients. According to the contractor responsible for the MRR, USAID is relying increasingly on technical assistance providers that serve lending institutions. USAID Has Identified and Disseminated Best Practices: USAID has funded several studies and projects, such as the Microenterprise Best Practices project, to publish emerging best practices. In addition, the agency has provided information on best practices to missions and implementing partners through policy guidance, training, and technical assistance. USAID has also collaborated with implementing organizations, microenterprise networks, and donors in disseminating information on best practices. Several organizations have published such information, including USAID; the Committee of Donor Agencies for Small Enterprise Development, whose secretariat is hosted and staffed by the World Bank; the Donor's Working Group on Financial Sector Development; Catholic Relief Services (CRS); and ACCION. (See app. II for a list of some key best practices.): USAID Has Made Efforts to Identify and Disseminate Best Practices: According to officials from the World Bank and other organizations, USAID has recognized the importance of identifying and disseminating successful and unsuccessful attempts to design and implement microenterprise activities. USAID's efforts include the following. * Growth and Equity through Microenterprise Investments and Institutions (GEMINI). Through its GEMINI project, which ended in 1995, USAID supported more than 120 studies on microenterprise development to publicize the experiences of leading microenterprise practitioners and experts in analyzing and managing microenterprise activities. These studies focused on the growth and dynamics of microenterprise in general and new approaches to delivering financial and nonfinancial assistance. The studies included data collection strategies for surveys; strategies for MFIs in Indonesia to more profitably provide financial services to the poor, strategies for MFIs to help microenterprises grow into small enterprises, recommended options to improve support for microenterprise development in Ecuador, and analyses of the importance of providing needed equipment to MFIs. * Microenterprise Best Practices project. USAID funded the Microenterprise Best Practices project, a research-oriented effort to develop and disseminate best practices. The project, completed in 2001, resulted in more than 100 reports, including concept papers, case studies, and technical tools and manuals providing guidance for designing and managing microenterprise activities. These reports included a model for Internet-based information for microenterprises in the Philippines, a description of Opportunity International's experience in Bulgaria and Russia in managing a microfinance program during a period of high inflation, a guide for reporting financial performance, and case studies of the difficulties encountered in converting nongovernmental institutions to commercial banks in Bolivia and Panama. * Guidance. USAID policy guidance for microenterprise development encourages missions to develop broad outreach activities to as many of the poor as possible, requires that MFIs charge unsubsidized interest rates to borrowers to cover the cost of operations, advises that missions consider and address host government policy constraints, and emphasizes the need for steady movement toward sustainability to achieve significant impact and institutional viability. Further, USAID policy states that missions providing assistance for microenterprise development should monitor and report on their outreach to the poor, including the distribution of its loan portfolio by loan amount. * Training. USAID supports training courses for its own and implementing organizations' staff in designing and executing microenterprise activities. In addition, USAID provides funding to the Microenterprise Development Institute at Southern New Hampshire University and to Business Development Services Training Program at the Springfield Centre in Durham, United Kingdom, to support their microenterprise development training programs. USAID's Accelerated Microenterprise Advancement Project provides scholarships to USAID staff for microenterprise development training and exchange programs. USAID also provides funding to the Small Enterprise Education and Promotion Network (SEEP), a network of nongovernmental organizations that implement microenterprise activities, for scholarships to enable USAID employees and practitioners to attend SEEP's training courses. During our fieldwork, we found that several USAID officials working on microenterprise development had received training at these locations. * Technical assistance. USAID's Prime Grant project provides direct technical assistance to missions in planning microenterprise activities. The project provides information on advances in microenterprise development and lessons learned from the missions' counterparts throughout the agency. USAID's Accelerated Microenterprise Advancement Project provides missions technical assistance from microenterprise experts and information on ongoing research and learning in microenterprise development. Missions may also receive technical assistance in developing scopes of work, including sample scopes, and ongoing support throughout the procurement process. USAID Has Collaborated with Other Organizations in Identifying and Promoting Best Practices: USAID has collaborated with implementing partners, networks of implementing organizations, and the World Bank in identifying and promoting best practices. These organizations have published handbooks, bulletins, and other documents on best practices; maintained Internet sites devoted to best practices; and sponsored seminars and workshops. For example: * CRS and ACCION International, two USAID-supported nongovernmental organizations, have published handbooks to assist in designing and implementing microenterprise activities. CRS also established the Microfinance Alliance for Global Impact project to help its implementing institutions strengthen their activities. The Foundation for International Community Assistance produced an evaluation of current practices used by microfinance institutions in assessing client poverty levels. Implementing partners such as ACCION International and Opportunity International also sponsor regional conferences, workshops, and seminars on best practices. * USAID has provided funding to SEEP to support its efforts to promote best practices. According to SEEP's Executive Director, USAID has been one of its leading supporters, providing funding and technical assistance and participating in the network's conferences and workshops. SEEP has published two studies on microenterprise best practices.[Footnote 32] The network also sponsors conferences and workshops on improving microenterprise activities. * USAID is a member of the World Bank's Committee of Donor Agencies for Small Enterprise Development and Donor's Working Group on Financial Sector Development, which have published guiding principles for designing and implementing microenterprise activities in 1995 and in 2001. The World Bank has also published reports to support the design and implementation of microenterprise activities, such as a handbook, to assist in operational planning and internal audits of activities.[Footnote 33] The group also provides information through MicroBanking Bulletin and the Microfinance Information Exchange. Conclusion: We found that microenterprise projects--including those funded by USAID--can help alleviate some of the impacts of poverty on individuals, households, and families. However, evidence suggests that microfinance alone has not lifted large numbers of the poor over the poverty line. In addition, despite USAID's use of micro loans to target the very poor, as mandated, few loans appear to be reaching this group, in part because loan size is an inadequate targeting method. Other evidence suggests that loans to the very poor can place some borrowers at risk of unmanageable debt and may be more beneficial when offered with other financial services such as savings and insurance and with development assistance such as grants, health services, education, and housing. Efforts to reach the very poor that do not recognize and address these key concerns may not be fully effective. Despite the general reliability of its data, certain methodological weaknesses in USAID's MRR system may prevent the agency from reporting with precision its program expenditures, the percentage of its funds going to the very poor, the percentage of MFIs that are sustainable, and the extent of USAID's contributions to the institutions it supports. Recommendation to USAID: We recommend that the Administrator of USAID review the agency's MRR system with the goal of ensuring that its annual reporting is complete and accurate. Specifically, the Administrator should review and reconsider the methodologies used for collection, analysis, and reporting of data on annual spending targets, outreach to the very poor, MFI sustainability, and the contribution of USAID funding to the institutions it supports. Agency Comments and Our Evaluation: USAID provided written comments on a draft of this report (see app. V). USAID concurred with the report's recommendation that it make improvements in its MRR. The agency cited three points with which it took issue, related to reaching the very poor, the sustainability of MFIs it supports, and its reporting of contributions to institutions. USAID stated that the number of small loans it had issued indicated that it was reaching the very poor. As discussed in our report and acknowledged in USAID's comments, loan size is now recognized as an inaccurate indicator of the extent to which this program is reaching the very poor. Given this limitation, we reviewed detailed impact studies that collected information on borrowers' economic status (see app. III for a summary of key studies on this topic); further verified this information through detailed discussions with international experts, USAID officials, and their implementing partners working with USAID-funded programs; and conducted detailed program reviews in three countries . The general consensus across the studies, experts, and program implementers is that microfinance projects serve those clustered around the poverty line but generally do not reach the very poor. USAID also stated that, contrary to our report, the agency uses a single definition of sustainability, and it inferred that the sustainability data reported in the MRR was accurate. We disagree with USAID on these points: We documented several definitions and interpretations that affect the reliability of the reported data, and we have added information to the report to clarify our concern regarding the agency's method for measuring microfinance institutions' sustainability. As noted in the report, 38 percent of MFIs that received USAID funding in fiscal year 2001 reported that they had achieved financial sustainability. The higher figure cited in USAID's response combined data on operational and financial sustainability, despite the fact that operational sustainability is, by USAID's definition, an interim measure toward the goal of achieving full financial sustainability. USAID stated that it would be difficult to allocate the microenterprise accomplishments reported in the MRR between USAID and other donors. However, it said that it plans to include more explicit language in the MRR to indicate that results are generally reported for entire institutions and that the resources of other donors and supporters contributed to the results. In its comments, USAID also agreed to (1) provide more explicit instructions on what activities to include in the MRR; (2) revise the formula for estimating the extent of funding that benefits the very poor and include in its annual report additional language concerning the formula; (3) improve the accuracy of data on obligations and poverty lending; and (4) adopt a new standardized definition of sustainability if one is adopted by the field. We believe that these improvements would be responsive to our recommendation and, if made, could improve the accuracy and balance of the MRR. We will send copies of this report to interested congressional committees as well as the USAID Administrator. We will also make copies available to others upon request. In addition, this report will be available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-3149 or at gootnickd@gao.gov. Other contacts and staff acknowledgments are listed in appendix VI. Sincerely yours, David B. Gootnick: Director, International Affairs and Trade: Signed by David B. Gootnick: [End of section] Appendix I: Scope and Methodology: To determine whether the U.S. Agency for International Development's (USAID) microenterprise development program is meeting its key objectives, we first identified those objectives by reviewing the agency's policy guidance and the pertinent legislation. We also held discussions with USAID officials in Washington, DC. To determine the results of the agency's microenterprise assistance, we met with officials and reviewed documents from USAID and their implementing partners in Washington, Peru, Egypt, and Bulgaria, and we met with program beneficiaries in these three countries. We selected Peru and Egypt because they have mature programs that have existed since the late 1980s and received high levels of USAID obligations over the past 10 years. These countries also represent culturally different areas (e.g., the program in Peru serves a large indigenous population, and the primarily business-led program in Egypt serves a combination of urban and rural areas). We selected Bulgaria because the program was relatively new; per capita income and the gross domestic product were high; and participants were reported by USAID to have higher educational levels and to be operating different types of businesses than in Africa, Asia, or Latin America. In addition, we reviewed a broad range of program and academic studies on the issues and conducted interviews and round-table discussions with academics and practitioners who have expertise regarding the ability of microenterprise activities to meet USAID's objectives. We also reviewed USAID studies that pertained to countries we visited, as well as studies that assessed project impact related to key program objectives. Because most available USAID data and most of the research literature focuses on microfinance, particularly micro loans, we concentrated our review primarily on this aspect of microenterprise development. To assess the reliability of the Microenterprise Results Reporting (MRR) data, we reviewed the survey questionnaires that are used to collect the data, noting strengths and weaknesses in the survey design. We also conducted a variety of analyses of the MRR database. Our analyses focused on the data on obligations supplied by the USAID missions and the data on microenterprise activities supplied by microfinance institutions (MFI), business development service (BDS) providers, and policy service providers from 1995 through 2002. We conducted interviews focused on data reliability with the contractor that manages the data collection and analysis and drafts the MRR reports. In these interviews, we asked how the survey data are collected, what quality checks are performed, and what other internal controls are in place. On our field trips to Peru, Egypt, and Bulgaria, we conducted data reliability interviews with officials at all three USAID missions and at six institutions that had received USAID funding. During our meetings with USAID missions and the institutions, we conducted spot checks of key MRR data to assess their reliability. We found that the reliability of the lending and BDS institutions' data on the percentage of women clients sufficed for our analysis, provided we noted that some BDS providers could not directly estimate these percentages. The data on lending institutions' sustainability were of uncertain reliability because of inconsistencies in the way respondents interpreted the MRR survey question; however, these data were consistent with the testimonial and documentary evidence that we gathered. To examine USAID's role in identifying and disseminating best practices, we reviewed (1) USAID policy guidance, (2) USAID country strategies and annual reports for the three countries we visited, and (3) other relevant USAID documents. We also reviewed a wide body of literature on the subject, including World Bank publications and the MicroBanking Bulletin; analyses of best practices produced by donor groups; handbooks, analyses, and other documents produced by USAID implementing organizations such as Catholic Relief Services and Opportunity International; and studies and analyses by recognized microenterprise experts. We interviewed USAID officials in the Microenterprise Development Division, the regional bureaus that oversee mission activities, and the countries we visited, including officers responsible for economic growth and microenterprise activities. We also interviewed officials of the World Bank and from implementing organizations in Washington, D.C; Baltimore, Maryland; and the countries we visited. Finally, we attended a roundtable on best practices whose members included recognized experts on microenterprise development from the World Bank, implementing organizations, and academia. The World Bank also provided informal comments on a draft of this report. We conducted our review from December 2002 through September 2003 in accordance with generally accepted government auditing standards. [End of section] Appendix II: Microenterprise Best Practices: Best practices are processes, practices, and systems that have been used by organizations and widely recognized as improving performance in achieving program goals. Although the research literature and our fieldwork indicate that no standard manual of best practices exists for microenterprise development, a core of preferable strategies (best practices) has emerged within the microenterprise industry comprising USAID, other donors, and their implementing partners. Design-Related Best Practices: * Perform due diligence reviews. USAID officials require their implementing partners to carefully review all candidates and to pay particular attention to choosing institutions with strong management skills. Officials from Catholic Relief Services (CRS), a nongovernmental organization that manages a microenterprise activity in Bulgaria, chose their implementing partner based on the partner's strong management experience. * Develop broad outreach. At USAID missions in Peru, Bulgaria, and Egypt, microenterprise activities included provisions for small loans to poor microentrepreneurs with no other affordable credit alternatives. A USAID-supported MFI in Egypt recently initiated a lending program that specifically provides small loans to the poor and is instituting a grant program to help the very poor become eligible for micro loans. In Peru, to target the poor and very poor, USAID chose to implement microenterprise activities in several of the country's poorest regions. USAID-supported institutions in Bulgaria and Egypt offered financial incentives to loan officers, based in part on the number of loans in their loan portfolio, to encourage them to attract clients. * Increase access to services. At implementing organizations in Peru, Bulgaria, and Egypt, loan officers assist poor and very poor clients in filling out the loan applications and attempt to review and approve loan requests within a few days. In addition, because the poor usually lack the collateral needed to qualify for loans, USAID supports collateral substitution activities to attract the poor and very poor who would have no other access to credit. USAID missions in Bulgaria, Egypt, and Peru conduct microenterprise activities that used group lending as a collateral substitute. For individual loans, an implementing organization in Bulgaria require that clients obtain written loan guarantees from acquaintances as a collateral substitute. * Adopt an appropriate lending model. Some models, such as group lending or village banking,[Footnote 34] may be more appropriate than individual lending programs for certain activities or institutions. CRS adopted a group lending model to serve the needs of the poor in Bulgaria. (This model also supports the CRS goal of advancing social and economic justice by serving the poorest.) A study of group lending activities in Africa, Asia, and Latin America indicated that more successful group lending models vary according to the local culture. For example, in South Africa, the South African Get Ahead Foundation adapted the traditional African rotating savings program to create similar group lending activities. * Offer an array of services. In addition to credit, services such as savings options and insurance are valuable to clients and, by providing other revenue sources, can assist MFIs in reaching sustainability. In Indonesia, a CRS MFI established a savings program in a village bank for its microenterprise clients. ACCION International, using a group lending model, has each member contribute a minimum amount to a common pool of savings. * Establish appropriate pricing policies for services. USAID requires financial institutions that receive microenterprise funding, even those that emphasize lending to the very poor, to charge unsubsidized interest rates. For example, CRS specifies in its microenterprise handbook that the MFIs it manages should charge unsubsidized interest rates. In Bulgaria, Egypt, and Peru, the annual interest rates can be as high as 40 percent annually, although the repayment period is often less than 1 year. * Control loan delinquency rates. Loan delinquency rates greater than 10 percent have been found to seriously undermine MFI sustainability. Several MFIs offer financial incentives to their loan officers partially based on the repayment rates of their loan portfolio. MFIs in Peru, Bulgaria, and Egypt use different methods to determine financial incentives to reward their loan officers, but all base the amount of financial incentive on the loan repayment rate of the officer's portfolio. At one implementing organization in Egypt, loan officers must maintain at least a 97 percent repayment rate on their loan portfolios to be eligible for financial incentives. An MFI in Bulgaria that provides individual loans requires that five people provide guarantees for each loan. The MFI also employs a loan collection officer and an attorney to file in court to collect on delinquent loans. MFIs in Egypt and Bulgaria that focus on poverty lending use a group lending model to provide for prompt loan repayment. * Address potential policy constraints. USAID guidance advises missions to consider the local economic environment when designing and implementing microenterprise activities. For example, the guidance advises missions not to provide assistance to MFIs during periods of high inflation. USAID officials told us the agency suspended a microenterprise activity because the Government of Egypt tax policies were too restrictive. Grant agreements may also include a component focused on policy and regulatory reforms to facilitate microenterprise activity. Such reforms may include permitting financial institutions to offer savings to clients, streamlining business registration procedures and assisting microentrepreneurs in registering and obtaining title to their businesses' assets. A USAID grant agreement in Bulgaria required the implementing organization to coordinate efforts in legislation on policy reform within 3 years. Implementation-Related Best Practices: * Require transparency and accountability in operations. USAID requires implementing partners and MFIs to report annually on financial and operational performance. In Peru, a USAID-funded technical assistance provider conducts audits of more than 40 MFIs to assess their implementation-related practices. * Provide adequate resources to successfully manage a microenterprise activity. Effective management information systems and other assets are necessary for implementing organizations and financial institutions to make decisions, motivate performance, and provide accountability over funds. USAID provides resources to help implementing partners and MFIs improve their management capacity. For example, most USAID grant agreements provide funding to rent office space, to purchase management information systems, including the computers needed to track outstanding loan balances and due dates, and to purchase other equipment such as office furniture. * Provide necessary training. Training for implementing organizations and clients in areas such as financial management and computers is often needed to ensure that MFIs manage operations effectively. USAID grant agreements may provide funding for training of implementing organizations' staff. Implementing organizations, such as an Opportunity International-funded MFI in Colombia, includes weekly training of clients in areas such as marketing and product presentation. A USAID-funded study of five nongovernmental organizations implementing microenterprise activities concluded that heavy investment in training was a factor in the success of village banks. * Provide incentives to loan officers. USAID-supported MFIs in Bulgaria and Egypt provide loan officers incentive-based salaries. Criteria for the incentives included the number of clients recruited and the clients' loan repayment rates. Incentives can double loan officers' monthly earnings. * Require a manageable loan portfolio for loan officers. Implementing organizations in Bulgaria and Egypt, such as CRS and Opportunity International, limit the number of clients one loan officer can manage. The number can vary depending on the ability of the officer, but few manage more than 300 clients. Also, because incentives are based primarily on loan default rates, officers are motivated to limit their pool of clients to a size they can manage effectively. * Monitor and evaluate performance. Donors and implementing organizations should monitor the performance of the MFIs they support to ensure they are meeting program goals. These goals can include a focus on the poor and women, as well as financial indicators, such as repayment rates and progress toward sustainability. They should also perform audits on a regular basis to ensure the accuracy of information reported. USAID typically performs midterm and final evaluations of its grant agreements. A cross-country study of village banks in seven countries concluded that oversight of the banks' operations was a critical factor in their success. Further, USAID requires annual financial reports of implementing organizations. * Promote MFI sustainability. This goal supports (1) sound financial practices, (2) expanding and maintaining outreach, and (3) reducing dependency on donor support. USAID and other implementing organizations encourage MFIs to charge unsubsidized interest rates to cover the cost of operations. A USAID-funded study of successful microenterprise activities in Indonesia and Bangladesh concluded that MFIs must have a commitment to, and a plan for, reaching sustainability.[Footnote 35] [End of section] Appendix III: Review of Microenterprise Studies: We collected, reviewed, and analyzed a set of 22 studies, 20 of which provide an overview of existing research and practice in microenterprise or its components. The purpose of our review was to obtain general information on the primary findings, issues, and debates in microfinance and microenterprise and to complement other USAID- specific components of our data collection efforts. We selected these studies on the basis of three criteria: (1) each study was published in 1998 or later, (2) each was peer reviewed or published by journals or publishers respected in the field, and (3) each was recommended by 2 or more of 15 microfinance experts we consulted. We also included two case studies (published and peer reviewed), because 5 or more of the experts we consulted recommended them. We ensured that the studies selected covered a range of microenterprise subtopics and scientific journals relevant to economic and social development issues. A primary reviewer summarized each study using a data collection instrument developed specifically for the purposes of this review. A secondary reviewer then verified each study summary. We reviewed each of the 22 articles and summarized the findings and relevant suggestions or recommendations, paying particular attention to information on poverty impacts, outreach, women's empowerment, sustainability, and best practices (see table 1). (Note: The recommendations in the third column of the table are those of the studies' authors and do not represent the viewpoint of GAO.): Table 1: Findings and Recommendations from 22 Selected Studies on Microfinance and Microenterprise Development: 1; Author title source: Bhatt, Nitin; Tang, Shui-Yan; Delivering Microfinance in Developing Countries: Controversies and Policy Perspectives; 2001. Policy Studies Journal 29 (2): 319-333; Primary findings: * Attempts to replicate best-practice models of microfinance based on successful programs (Grameen, BRI, BancoSol) have largely been disappointing; * Few programs have lived up to their original objective of "including the excluded."; * In most cases, subsidies have ended up funding inefficient and lax management practices that have resulted in limited outreach, high loan default rates, and unsustainable operations; Author recommendations: * The best solution for the microfinance industry is to have a diverse array of institutions that are both non-and for-profit, subsidized and unsubsidized, and bare-bones and integrated; * There is a need for better, more accurate financial reporting and cost-benefit analysis in microfinance. 2; Author title source: Cohen, Monique; Making Microfinance More Client-Led; 2002. Journal of International Development 14 (3): 335- 50; Primary findings: * The microfinance industry has been largely a single product industry; this product has worked well for people around the poverty line but is less effective at reaching the very poor; * Although most MFIs serve a wide range of clients, the majority of clients are clustered just above and just below the poverty line; * Many poor people are highly indebted and microfinance loans are only one component of the debt burden of many households; * High drop-out rates in some places have arisen as competition between MFIs has increased; * Poverty-targeted programs tend to reach a higher percentage of lower income clients; even here, poorer populations often exclude themselves from microfinance programs; * Many poor people "patch" together funds from a variety of sources, including MFIs, in order to meet their consumption and investment needs; * Most MFI management information systems lack adequate client information that can be used as a tool in decision making and product development; Author recommendations: * To achieve outreach to underserved market segments and to the very poor, the microfinance industry must focus on clients' specific financial needs and design appropriate financial services accordingly. 3; Author title source: Steven Haggblade; Mead, Donald C; Meyer, Richard; An Overview of Policies and Programs for Promoting Growth of the Rural Nonfarm Economy; Forthcoming. In Developing the Rural Nonfarm Economy, edited by Steve Haggblade, Peter Hazel, and Thomas Reardon; Primary findings: * In analyses of the impacts of micro loan programs on rural nonfarm economies, outreach has proven impressive, financial sustainability has proven elusive, and income increases have been generally documented at the household and firm levels, particularly for the better-off poor and in Asia; * Micro loans have not advanced our understanding of the challenge of lending to enterprises in rural areas where population densities are lower and where the seasonality in cash flows and lending costs is higher; * Business development services involving the delivery of nonfinancial services to rural nonfarm enterprises are more heterogeneous, more problematic, and have produced decidedly fewer results than the delivery of financial services to rural nonfarm enterprises; Author recommendations: * In the future, improvements in rural welfare will need to go beyond financial markets and services to develop new engines of economic growth, new technologies, and new ways of doing business; * A focus on subsectors, supply chains, and clusters can lead to tailored systemic interventions that unleash growth potential for large numbers of rural firms. 4; Author title source: Hulme, David; Impact Assessment Methodologies for Microfinance: Theory, Experience, and Better Practice; 2000. World Development 28 (1): 79-98; Primary findings: * Microfinance impact assessments vary greatly in their range, rigor, and practicality; there have been relatively few quasi-experimental, control-treatment studies, and practitioners have found such studies to be expensive and difficult to use; * There is increasing movement toward mixed studies incorporating quantitative and qualitative elements and toward designing impact assessment activities into microfinance programs; Author recommendations: * Impact assessments must be tailored to the resources, needs, and time frames defined by the program's context. 5; Author title source: Khandker, Shahidur R; Fighting Poverty with Microcredit: Experience in Bangladesh; 1998. New York: Oxford University Press; Primary findings: * Globally, micro loan programs have been able to reach the poor and enhance both their productive and human capital by generating self-employment; * The long-run cost- effectiveness of micro loans depends on the overall growth of the economy; * Bangladesh is one of the few countries in which micro loan programs have been successfully replicated; * Microfinance in the Bangladesh programs reduced poverty by increasing per capita consumption among program participants and their families; poverty reduction estimates based on consumption impacts show that about 5 percent of program participants can lift their families out of poverty each year through microfinance; * One percent of rural households in Bangladesh can free themselves from poverty each year through micro loans; * Participation in micro loan programs enhanced women's productive means by increasing their access to cash income from market- oriented activities and by increasing their ownership of nonland assets; * Subsidies in the Bangladesh programs studied were necessary to defray an array of high costs associated with providing microloans to the poor; Author recommendations: * In replicating microfinance programs, it is necessary to design a socially conscious and transparent system of accountability that works for both program officials and borrowers; * The poor and women need special banking, so the commercialization of micro loans should not divert attention from meeting this special need. 6; Author title source: Matin, Imran; Hulme, David; Rutherford, Stuart; Finance for the Poor: from Microcredit to Microfinancial Services; 2002. Journal of International Development 14 (2): 273-94; Primary findings: * Poor people must and do save in order to finance the large lump sum payments needed to deal with an array of life-and environment-based needs and crises; * The poor are a heterogeneous group of vulnerable households with complex livelihoods and varied needs; * The success of the group lending model globally is mixed; * Weekly frequency of repayment and repayment incentives (typically larger repeat loans) correlate strongly with success, defined in terms of arrears rate and subsidy repayment index; Author recommendations: * A shift away from credit for business development toward financial services of all types for the poor will help them to effectively manage the many risks and crises that emerge in their lives. 7; Author title source: Mayoux, Linda; From Access to Empowerment: Widening the Debate on Gender and Sustainable Micro-Finance; 2000. Journal fur Entwicklungspolitik 16 (3): 247-273; Primary findings: * Some policies introduced for financial sustainability result in adverse impacts on empowerment; * Beneficial impacts of microfinance per se on women cannot be assumed; impacts vary substantially across cases and programs; * Microfinance services have enabled some, but not all, women to increase their incomes and negotiate improvements in their family and community position; microfinance alone is the most limited for the poorest and most disadvantaged women; * Women's decisions about loan use involve assessments of benefits, costs, and risks in the context of the gender norms of household resource allocation and decision-making; * Women's choices about activity and ability to increase income are constrained by gender inequality; market, resource, and skill constraints; and microfinance delivery; Author recommendations: * An empowering gender policy should include increasing women's incomes from their own activities; control of income from loans and activities; negotiated improvements in household well- being, and access to support networks; * A gender policy that supports financial sustainability and empowerment should consider improving the conditions of service delivery, reducing costs of complementary services (e.g., training, gender awareness/support), and mainstreaming gender policy. 8; Author title source: Mead, Donald C; Liedholm, Carl; The Dynamics of Micro and Small Enterprises in Developing Countries; 1998. World Development 26 (1): 61-74; Primary findings: * House-to-house baseline surveys make clear that the number of micro and small enterprises (MSE) is far larger than that reported in most official statistics; * In most countries, the majority of MSEs operate in rural areas, are engaged in the trade and manufacturing sectors, are owned and operated by women, and are more likely to be operated from the home; * Younger, smaller, male-headed manufacturing and services MSEs in urban areas are more likely to grow than are older, larger, female-headed trade MSEs in rural areas; * When the economy is more buoyant, new jobs come from the expansion of existing firms; in times of stagnation, existing MSEs cut back on employment and a larger percentage of new jobs comes from start-ups; Author recommendations: * Policies and projects for supporting MSE development must be aware of the diversity of MSEs and focus on the types of enterprises and on the stages in the enterprise's life cycle where interventions can do the most good. 9; Author title source: Morduch, Jonathan; The Microfinance Promise; 1999. Journal of Economic Literature (37) 4: 1569-1614; Primary findings: * There are few reliable studies of the net poverty impacts of microfinance programs; those which do exist suggest some limited positive impacts, as well as other mixed results; * Sustainable microfinance programs have by and large not been able to reach the very poor in large proportions or numbers; * Microcredit supplements clients' income and does not produce fundamental shifts in clients' employment patterns; microcredit rarely generate new jobs for others; * The features of microfinance program designs vary to a high degree globally; * Incentives and specific program design elements such as group lending, weekly repayments, graduated repayment schedules, and forced savings, have been important in the successes that microfinance programs have achieved to date; * Microcredit has had very limited success in regions with low population densities and with seasonal income patterns; Author recommendations: * Savings programs will strongly complement the credit-driven approach that has dominated microfinance to date; * To alleviate poverty, new approaches to management structures and product/program design are necessary. 10; Author title source: Morduch, Jonathan; The Microfinance Schism; 2000, World Development 28 (4): 617-629; Primary findings: * There is a strong consensus among prominent policy making organizations that microfinance institutions (MFI) that focus on profitability (through high interest rates) will have the strongest effects on poverty; * This consensus on best practices is at odds with actual practice, where the large majority of MFIs are not focused on profitability, have not given up subsidies, and are not sustainable; * Moderate subsidies have been necessary in some programs to attain social objectives not present in for-profit MFIs; * Efficient operations and clear incentive structures for clients and loan staff are more important factors in achieving success in microfinance than a focus on profitability; nonprofit, subsidized programs have been able to operate efficiently, provided they adhere to clear budgeting and performance criteria; * Governments have played critical indirect roles in cases such as Indonesia's BRI and Thailand's BAAC, which are state-owned commercial banks; Author recommendations: * A diverse array of socially oriented, subsidized, and profit-oriented, sustainable microfinance institutions is best for the microfinance industry and for wide-scale poverty alleviation; * If subsidized programs are to continue at current funding levels, they will likely need to rely increasingly on their own governments. 11; Author title source: Navajas, Sergio; Schreiner, Mark; Meyer, Richard; Gonzalez-Vega, Claudio; Rodriquez-Meza, Jorge; Microcredit and the Poorest of the Poor: Theory and Evidence from Bolivia; 2000. World Development 28 (2): 333-346; Primary findings: * In a study in La Paz, Bolivia microfinance lenders tended to serve those near the poverty line, not the "poorest of the poor."; * Rural and group lenders tended to serve larger numbers of poorer clients than urban and individual lenders; * In addition to depth, defined as the degree to which an MFI reaches the very poor, factors such as worth, cost, breadth, length, and scope are important in evaluating outreach to and the social welfare of the poor; * Length of outreach (time frame in which an organization produces loans) is the key aspect of outreach because it prompts improvements in the five other aspects; Author recommendations: * Policy-makers and MFIs should consider all aspects of outreach in making judgments about the focus and social benefits of microfinance. 12; Author title source: Rankin, Katherine N; Social Capital, Microfinance, and the Politics of Development; 2002. Feminist Economics 8 (1): 1-24; Primary findings: * There is substantial evidence that microfinance and micro loan programs have served to strengthen existing patterns of male and class-based power relations, rather than empowering women; Author recommendations: * Development agencies and policy should place greater emphasis on finding ways to empower the poor and disadvantaged so that they can overtly challenge dominant social power relations. 13; Author title source: Rhyne, Elisabeth; Microfinance Institutions in Competitive Conditions; 2002. In The Commercialization of Microfinance: Balancing Business and Development, edited by Elisabeth Rhyne and Deborah Drake. Kumarian Press; Primary findings: * The microfinance industry has to date primarily been a one-product industry, based on short-term (3 to l2 months) group-based lending, with frequent repayments, and small loan sizes; * Microfinance institutions have traditionally taken an inward-looking approach to their development, focusing primarily on streamlining operations to achieve greater outreach; * In some countries, there has been a substantial increase in competition among MFIs for clients, e.g., Bolivia, Bangladesh, Nicaragua, and Uganda, which has increased the need for MFIs to be more oriented toward their 1) competition and 2) clients; * Overlending is a key problem that arises in competitive markets; Author recommendations: * Maintaining sustainability, while becoming more attuned to the needs of clients, is critical; MFIs must adapt to the changing needs of clients to survive and develop in the newly competitive microfinance environment; * MFIs must be attuned to the problem of overlending. 14; Author title source: Robinson, Marguerite S; The Microfinance Revolution: Sustainable Finance for the Poor (Chapters 1 and 2); 2001. Washington, D.C.: The World Bank; Primary findings: * Sustainable microfinance only occurs in systems providing commercial financial intermediation; other models (e.g., subsidized credit) are not sustainable in the long term and are not affordable on a global scale; * Only institutional commercial microfinance can provide sustainable financial services to the working poor by providing low-cost credit and wide outreach; * The lowest levels of the poor need food, employment, and/or government or donor assistance and grants; Author recommendations: * The shift away from donor-assisted credit delivery to sustainable financial institutions that provide commercial microfinance is essential to meet the global demands for microfinance; * Governments must support commercial microfinance through regulation, supervision, and public education to ensure success. 15; Author title source: Rutherford, Stuart; The Poor and Their Money; 2000. New Delhi, India: Oxford University Press; Primary findings: * The poor can save, do save, and want to save money; the poor have to save to make the large lump sum payments that come up frequently in their lives; * The argument about whether the poor need savings or loans is false; they need both; * Financial services for the poor help them trade their savings for lump sums of cash; good financial services for the poor are those that perform this trade well; Author recommendations: * The key to microfinance is to find creative ways to collect small sums (savings, repayments, insurance premiums) and turn them into large sums (loans, withdrawals from savings, or insurance pay-outs). 16; Author title source: Schreiner, Mark; Aspects of Outreach: A Framework for Discussion of the Social Benefits of Microfinance; 2002. Journal of International Development 14 (5): 591-603; Primary findings: * There are six interacting dimensions of outreach: worth to clients, cost to clients, depth, breadth, scope, and length; * Tradeoffs exist in achieving outreach: programs focusing on sustainability achieve breadth, length, and scope at the expense of depth; programs focused on poverty alleviation achieve depth at the expense of breadth, length, and scope; Author recommendations: * Microfinance institutions should focus on sustainability because the longer an MFI is in operation the greater the number of clients and the greater the overall impact on poverty. 17; Author title source: Sebstad, Jennefer; Cohen, Monique; Microfinance, Risk Management, and Poverty; 2000. AIMS. U.S. Agency for International Development; Primary findings: * Microfinance programs have been more successful in reaching clients from moderately poor and vulnerable non-poor households than from extremely poor households; * Clients use loans for a wide range of purposes beyond enterprise development; * Micro loans are more useful for clients in protecting against risks ahead of time than in smoothing consumption following a shock; * Investment in productive assets is more prevalent among clients having larger loan sizes, higher numbers of loans, and more time in the program; * Microloans have positive impacts on reducing vulnerability measured as diversified income sources, increased numbers and types of assets, and women's empowerment; Author recommendations: * Future programs should link services to clients' needs and demands, should become more flexible, and should diversify services offered; * Linking financial services with other supportive services could improve clients' ability to deal with risks. 18; Author title source: Snodgrass, Donald R; Sebstad, Jennefer; Clients in Context: The Impacts of Microfinance in Three Countries. Synthesis Report; 2002. AIMS. U.S. Agency for International Development; Primary findings: * Overall, microfinance makes a difference to people's lives; however, microfinance does not have the unequivocally large, positive impact on microenterprise development and poverty reduction that has been claimed for it by some; * Microfinance impact is a function of product characteristics, the macroeconomic situation, and the length of time the client has been in the program; * The microfinance programs reviewed had significant impact on a variety of individual-, household-, and enterprise-level income and well-being indicators, although the nature and magnitude of these impacts vary across countries and programs; * Microfinance provided recipients with resources for dealing with risks typically not available to nonrecipients; * Receiving a micro-loan resulted in greater household and enterprise decision-making authority for women in the study; Author recommendations: * Expanding and diversifying services to include savings, insurance, and emergency loans will increase impact and enhance clients' ability to cope with shocks. 19; Author title source: Snow, Douglas R; Buss, Terry F; Development and the Role of Microcredit; 2001. Policy Studies Journal 29 (2): 296- 307; Primary findings: * We have very scant knowledge of the relationships between program designs and outcomes; * There have been very few strong microfinance program outcomes studies; as a consequence, "there is good reason to fear reliance on program design as a surrogate for outcomes."; * If forced to stand alone, few micro loan programs, if any, would survive; * The twin concepts of sustainability and outreach are inherently contradictory; Author recommendations: * Governments, if not all donors, ought to have specific goals for micro loan programs and these should be grounded in strong research and knowledge of impacts. 20; Author title source: Von Pischke, J.D; Current Foundations of Microfinance Best Practices in Non-Industrialized Countries; 2001. In Replicating Microfinance in the United States, edited by Jim Carr and Zhong Yi Tong. Washington, D.C.: Fannie Mae Foundation; Primary findings: * Major players in both sustainability-and poverty-oriented MFIs have developed a consensus about preferred microfinance strategies; * They have generally agreed upon 12 best practices:; * Create sustainable institutions; * Charge interest rates that will enable the lender to cover costs; * Control arrears; * Use subsidies for institution building, i.e., technical assistance and capitalization; * Behave competitively from the outset; * Maintain uncompromising commitment to the target group; * Select "owners" who will provide effective governance; * Manage risk, explore failure; * Develop good management information systems; * Be transparent; * Focus on incentives; * Exchange information on defaulters; Author recommendations: [Empty]. 21; Author title source: Woller, Gary M; Dunford, Christopher; Woodworth, Warner; Where to Microfinance?; 1999. International Journal of Economic Development 1 (1): 29-64; Primary findings: * There is general agreement in microfinance that poverty reduction and the achievement of financial self-sufficiency are the basic goals for the industry; * The industry disagrees on how to achieve these goals: the institutionists emphasize financial self-sufficiency over a focus on the very poor; the welfarists emphasize assisting the very poor over the need for financial self-sufficiency; * The institutionist perspective has become more prominent in the microfinance industry, but its assertions that subsidized institutions cannot survive for long periods and cannot have strong, large-scale impacts on the very poor are not fully supported by existing evidence and are based on an array of questionable or inconsistent assumptions; * Most microfinance institutions lie somewhere along a continuum ranging from traditional business and traditional social service orientation; Author recommendations: * There is no need for a once-and-for-all choice between competing approaches; the microfinance industry should be characterized by a diversity of microfinance institutions that cater to various segments of low-income communities. 22; Author title source: Wright, Graham; Replication. Regressive Reproduction or Progressive Evolution?; 2000. Journal of Microfinance 2 (2): 61-81; Primary findings: * Replication of a best practices model for microfinance program design has been difficult, because the contexts in which such a model would be implemented vary so greatly; * Institutions attempting to replicate the same general microfinance model are characterized by a high level of diversity in implementation methodologies; * The poor often participate in a variety of formal and informal financial institutions; Author recommendations: * To be successful, microfinance programs must be adapted and tailored to the cultural, economic, social, and political features of the contexts in which they are implemented; * Implementation and replication of programs must include a period of research, experimentation, and adaptation. [End of table] [End of section] Appendix IV: USAID's Microenterprise Results Reporting: To monitor and evaluate its microenterprise portfolios, USAID developed a data collection process and information management system known as the Microenterprise Results Reporting (MRR). The term also refers to an annual report that presents the agency's financial data--primarily amounts it obligates for microenterprise development--and programmatic information. The MRR data are collected through annual surveys of USAID staff in headquarters and at overseas missions and the institutions that receive USAID funding. A USAID contractor is responsible for data collection and the management information system. Beginning early in each fiscal year, the contractor requests obligations data for microenterprise projects from USAID headquarters and missions. The mission staff report current year obligations and identify the recipient institutions, categorizing them as microfinance, business development services (BDS), or policy services providers. In addition, the mission staff identify institutions that received obligations in previous years for ongoing projects. Separate surveys have been designed for the microfinance institutions (MFI), BDS providers, and policy service providers. The survey for MFIs asks about outstanding loan balances, the number of loans to women, maximum loan sizes, loan loss, loans below the poverty lending threshold percentage, the number of rural clients, savings, and the financial sustainability of the institutions. The survey for BDS providers asks about the types of services provided, the number of clients overall, the number of women clients, the number of rural clients, the number of clients with "poverty loans," data sources for clients, the clients' industrial sector, the institutions' competitors, the demand for BDS, and exit strategies. The policy service providers survey asks about the types of institutions and for descriptions of policy issues covered. The number of respondents to the annual surveys during 1998 to 2001 has remained fairly constant, ranging from 361 to 411. Most MRR respondents complete the MFI survey. In 2000, for example, 512 surveys were sent out; 282 of the 361 respondents completed the MFI survey, 99 completed the BDS survey, and 18 completed the policy survey. The reported response rates rose in recent years, from 56 percent (411 surveys) in 1998 to 84 percent (492 surveys) in 2001. USAID contractor staff analyze the data and, in some cases, apply methodologies the agency has developed to assess whether it has met particular program goals, such as its poverty-lending target. This methodology is designed to weight the individual institutions' obligations by the amounts of loans that are considered poverty loans. The data and the analyses are presented in the annual reports, which also provide examples of USAID-funded microenterprise projects. In addition to publishing the data in the MRR reports, the contractor also publishes selected data on a Web site accessible to the agency's missions, institutions that receive USAID-funding, and interested others. [End of section] Appendix V: Comments from the U.S. Agency for International Development: Note: GAO comments supplementing those in the report text appear at the end of this appendix. USAID: U.S. AGENCY FOR INTERNATIONAL DEVELOPMENT: NOV 6 2003: Mr. David Gootnick: Director, International Affairs: U.S. General Accounting Office: 441 G Street, N.W.: Washington, DC 20548: Dear Mr. Gootnick: I am pleased to provide the U.S. Agency for International Development's (USAID's) formal response on the draft GAO report entitled "MICROENTERPRISE DEVELOPMENT: USAID's Program has Met Some Goals; Annual Reporting Has Limitations" (November 2003). We appreciate that the GAO has taken USAID's management of this important work seriously. The draft report's finding that we have met some key goals confirms the value of our comprehensive microenterprise strategy and the dedication of our staff around the world to this important program. The findings that our microenterprise development programs alleviate poverty and benefit poor entrepreneurs and their households are consistent with our experience of over two decades. The recognition of the strong and effective priority USAID places on serving women - who comprise two-thirds of all clients of USAID-supported microfinance institutions - is also welcome. As the draft report notes, USAID has focused consistently on the importance of building sustainable institutions, and this has paid off in an unusually high complement of sustainable partners. We are pleased with the findings concerning our contributions to promoting donor collaboration and generating and disseminating knowledge and best practices. We continue to find, however, that the report's primary focus on microcredit, and to some extent microfinance, is inconsistent with the draft report's title and does not address the full range and scope of USAID's microenterprise strategy and programs. In addition, we note three specific points made by the draft report which we consider to present errors of fact and interpretation and to result in inaccurate evaluation of USAID's performance 1. Reaching the Very Poor: The report's assertions that "USAID's Microfinance Activities Generally Have Not Reached the Very Poor" and "Small Loan Size Does Not Accurately Target the Very Poor" are contradicted by the only available data of which we are aware. Congress, USAID and the practitioner community have long used loan size as a proxy for service to the very poor; passage of the Microenterprise for Self-Reliance Act in 2000 made use of this proxy a statutory requirement. While recognizing that the loan size proxy was imperfect, USAID has applied it through its Microenterprise Results Reporting (MRR system) to analyze and report on service to the very poor. Using this measure, MRR reported that in FY 2001 (the most recent year for which completed data are available), USAID-supported MFIs made 2 million poverty loans out of 3 million total loans. Thus, two-thirds of all loans met the statutory standard of service to the very poor. While the use of smaller loans is one way to increase the likelihood of the very poor benefiting, USAID has long supported a wide variety of methods to target services to the very poor, including the use of collateral substitutes, service features that overcome cultural barriers to access, and safe and convenient savings services. 2. Sustainability Performance and Measurement: The report's finding that "Data on MFI Sustainability Are Not Standardized" is not correct. For more than five years, MRR has used a single, clear definition of Sustainability (distinguishing between operational and financial sustainability) in the questionnaires distributed to fund recipients. There is currently no single alternate definition of Sustainability that has been adopted as an industry standard. USAID's microfinance partners have demonstrated impressive sustainability; at both levels in FY 2001, over half were sustainable and, more important, 73% of clients were served by sustainable institutions. 3. Joint Funding of Microenterprise Programs: The report voices concern that USAID's annual MRR reports do not isolate the Agency's contributions to the MED institutions it supports, making it difficult to determine the impact of the agency's funding. The reports do not attempt to disaggregate USAID's impact from that of other donors, investors, lenders, or the institution's earnings and assets, for two important reasons. First, it would be impractical and methodologically questionable to allocate impact between USAID and other supporters, since we typically fund institution-building rather than discrete, separable projects. Second, Agency policy and philosophy is oriented toward engaging others in support of our implementing partners in all fields, including microenterprise development. Such efforts would be seriously hindered by actions on our part that appeared to take credit for results that are the product of joint support. GAO Recommendation: Corrections of these points would, we feel, add to the credibility of the report. Consequently, we find the recommendation to be reasonable. It should be noted that 2003 amendments to the Microenterprise for Self-Reliance Act also require substantial changes to the MRR and reporting by awardees on the extent of their service to very poor clients. The Microenterprise Development Team of the Office of Poverty Reduction, plans to address the areas of concern within the next month, with a view to presenting the Administrator the proposed MRR changes by mid-November. The Administrator's decisions on these matters will be communicated to you as soon as possible. It should be noted that the MRR has already incorporated a suggestion that came from the GAO team prior to receipt of the draft report, i.e., that USAID missions receive more explicit instructions on how to pro- rate reporting of activities that serve purposes and target groups beyond microenterprise development and poor microentrepreneurs. By the date of publication of the GAO report, the Microenterprise Development team will have revised the formula used for estimating the extent of funding that benefits very poor clients in order to improve the accuracy of data on obligations and poverty lending; and will have drafted additional language for inclusion in the annual report concerning the estimating methodology. We do not believe there are significant weaknesses in the definitions or techniques MRR employs to measure sustainability of USAID-supported microfinance institutions, but we will review this with colleagues in other organizations. Further, if the microfinance field adopts a new standardized definition of sustainability, we will incorporate it into MRR. Similarly, because USAID has focused on institution-building, and has strongly encouraged partners to leverage other funding and support, in most instances we do not consider it practical or appropriate to analyze and report the impact of USAID funds in isolation from those of other donors, investors, or creditors. However, we will include more explicit language in the MRR annual report indicating that results reported generally cover the entire institution and that the resources of other donors and supporters contribute to those results. No data system is 100% accurate, of course, and any system can be improved. We believe that these changes provide a reasonable and cost-effective response to the draft report's recommendation that will not significantly increase the burden on MRR reporters. Enclosed is Annex 1, which provides more detailed commentary on the substance of the draft report and USAID's response to the recommendation. Thank you for the opportunity to respond to the GAO draft report and for the courtesies extended by your staff in the conduct of this review. Sincerely, John Marshall: Assistant Administrator: Bureau for Management: Signed by John Marshall: Enclosure:; a/s; Annex 1: USAID Substantive Comments on Draft Report: We are delighted that the GAO has taken USAID's management of its microenterprise development work so seriously. The draft report's finding that we have met some key goals confirms the value of our comprehensive microenterprise strategy and the dedication of our staff around the world to this important program. The findings that our microenterprise development programs alleviate poverty and benefit poor entrepreneurs and their households are consistent with our experience of over two decades. The recognition of the strong and effective priority USAID places on serving women - which comprise two-thirds of all clients of USAID-supported microfinance institutions - is also welcome. As the draft report notes, USAID has focused consistently on the importance of building sustainable institutions, and this has paid off in an unusually high complement of sustainable partners. We are gratified with the findings concerning our contributions to promoting donor collaboration and generating and disseminating knowledge and best practices. We offer the following comments on the draft report. They are intended to address errors of fact or interpretation, and to explain USAID policy and practices relevant to the specific findings. USAID Programs do not Reach the Very Poor/Small Loan Size Does Not Accurately Target the Very poor: The draft report concludes that "microfinance generally has served the poor clustered around the poverty line but not the very poor" (page 1) and that "few loans appear to be reaching the very poor" (page 4). This conclusion is repeated at numerous points throughout the draft report. USAID does not believe that the evidence presented in the draft report supports this finding. * For well over a decade, Congress, USAID, and MED implementing partners have agreed to use loan size as a proxy for service to the very poor. * When the Microenterprise for Self-Reliance Act of 2000 came into effect, it created the statutory requirement that the Agency use this measure to analyze and report on the extent of service to very poor people (defined by the law as those at or below half of national poverty lines). * Using this measure, MRR reported that in FY 2001, USAID-supported MFIs made 2 million[NOTE 1] poverty loans out of 3 million total loans. In other words, two-thirds of all loans met the statutory standard of service to the very poor. USAID has long recognized the many challenges of ensuring that the very poor benefit from microenterprise assistance programs and has long supported a variety of methods to help ensure that the very poor are served with MED services. While the use of smaller loans is one way to increase the likelihood of the very poor benefiting, even in its 1995 policy paper, USAID suggested a number of other techniques such as use of collateral substitutes, unsubsidized interest rates, service features that overcome cultural barriers to access, safe and convenient savings services, etc. (USAID policy paper pages 14-15). As required by law, USAID has used the delivery of small loans as the proxy for measuring the extent of service to very poor clients, not as the means for targeting them. Loans to Very Poor May Be Ineffective and Can Have Negative Consequences: The draft report states that loans may affect poor borrowers more negatively than positively. USAID is well aware of the dangers of over- indebtedness for poorer clients and believes that the available empirical evidence does not support the draft report's suggestion that this is a significant or widespread problem. * The weight of empirical evidence and program experience suggests that very poor people typically benefit from access to well-tailored credit and other microfinance services. We do agree with the statement in the draft report that it is important to provide the poor, including the very poor, with an expanded range of financial products. * USAID is actively supporting research, pilot programs and expansion of successful efforts related to a range of products such as savings, insurance, and money transfer services. Some Progress Made Toward MFI Sustainability: The draft report concludes that USAID has emphasized the importance of the microfinance institutions (MFIs) it supports achieving sustainability. The sustainability track record of USAID-supported MFIs is very strong. Although the draft report cites the estimate that of roughly 10,000 MFIs worldwide, only 75-250 are estimated to be sustainable, * More than 50% of USAID-funded MFIs have achieved operational sustainability (where the MFI is covering all operating costs, including loan losses, from operating revenues) or the higher standard of full financial sustainability (where the MFI is also covering inflation, commercial costs of borrowing to finance expansion, and the imputed costs of granted capital). * In FY 2001, 53% of MFIs reporting to MRR were either operationally or financially sustainable. We consider this a remarkable achievement. In the sections covering MRR's methodology and reliability, the draft report asserts that MRR has not used a standardized definition of sustainability. This is not correct. For more than five years, MRR has used a single, clear definition of sustainability (distinguishing between operational and full financial sustainability), in part so as to be able to present comparable data from year to year. The two-part definition is included in the standardized questionnaire that goes to each recipient of microfinance funding, and the MRR contractors perform data quality checks to ensure consistent reporting. Poverty Lending Requirement May Hamper MFI Sustainability: The draft report suggests that sustainability might not be consistent with serving very poor clients. In particular, it suggests that poverty lending requirements may hinder MFI sustainability. While it would appear that sustainability is more difficult to achieve for MFIs reaching very deep with their services (primarily due to the smaller loan size and perception that the clientele is riskier), * USAID and others have found that even MFIs serving a relatively large share of very poor clients can achieve sustainability. * It is USAID's policy that even institutions that target the very poor should commit to achieving sustainability, although we recognize that this might require a longer time-frame. Basic MRR Data are Generally Reliable: The draft report found that the basic program data in USAID's Microenterprise Results Reporting system (MRR) are generally reliable. It found that the MRR data generally met the criteria of accuracy, completeness, and consistency. We agree and are proud that MRR is widely recognized as by far the most comprehensive and accurate donor reporting system and data base among microfinance and microenterprise donors. It is important to note that MRR is tasked with collecting, analyzing, and reporting data on USAID microenterprise obligations and the activities and results of its implementing partners. It is not intended to track microenterprise expenditures nor is it intended to perform internal audits on financial and programmatic performance. This is the responsibility of missions and other USAID operating units, which carry out these functions as part of their Performance Monitoring Plans and related data quality assessments. * A recent GAO audit of the Egypt mission's microfinance portfolio and operations, for example, found strong compliance with USAID policy guidance, strong monitoring systems and a high level of accuracy of performance data reported. Reliability of Obligations Data is Uncertain: The draft report raised questions about the reliability of obligations data reported by MRR. The main concern described by the GAO team focused on funding obligations that did not appear to solely serve microentrepreneurs or provide micro loans. We believe these concerns might derive from a misunderstanding of USAID policy and practice, as well as mission procedures in reporting obligations data to MRR. By policy, USAID does not require that MED awardees solely serve micro- scale entrepreneurs. For example, the microenterprise policy paper states: "USAID's concern is that organizations receiving its support maintain an effective commitment to include the poor in their service delivery, not that they exclude other groups" (page 16). * Serving a mix of clients (for example, small and medium firms as well as microenterprises, or households that include the very poor, moderate poor, and vulnerable non-poor) may result in stronger impacts on the target populations or firms, as well as improved prospects for sustainability. * By offering diverse credit and savings services to both very poor and less poor clients, for example, many credit unions reach far more very poor households than smaller MFIs that solely serve very poor households. * In "mixed" programs, the key concern is not that service be restricted to micro-scale firms; rather the funding reported to MRR must report only that portion that serves microenterprises. USAID missions are accustomed to pro-rating their obligations (e.g., in reporting obligations against various earmarks and directives), and staff and the MRR contractors showed evidence to the GAO team that such pro-rating of obligations occurs for MED activities. To respond to the concern articulated by the GAO team, however, MRR has recently included more explicit pro-rating instructions in its website guidance to missions and will also include additional language in its annual request for MED obligations. MRR Does Not Link USAID Funding to Services Provided: The draft report voiced concern that the agency's annual reports do not isolate USAID's contributions to the MED institutions it supports, making it difficult to determine the impact of the agency's funding . There are two important reasons that USAID does not attempt to disaggregate the impact of its funding from that of other donors, investors, and lenders. First, it would be impractical and very difficult to devise a reasonable methodology for doing so, since USAID is typically providing funding to build MED institutions, rather than for discrete projects with separable budgets, activities, and impacts. * There is a strong consensus in the microfinance donor community that it is not meaningful or cost-effective to try to impute impact among different donors and investors. The second reason that USAID does not isolate its contributions from others is more a matter of policy and philosophy. The agency is committed to a very active effort to engage others in support of our: implementing partners in all fields, including microenterprise development. The Global Development Alliance provides one visible example of this commitment to leveraging other resources, which we believe is inconsistent with highlighting the USAID contributions. In response to the concerns raised in the draft report, however, we will in the future include language in the annual report indicating that many awardees receive support from other sources as well and that those sources deserve a share of the credit for the awardees' impacts. It should also be noted that the draft report found that USAID has identified and disseminated best practices and cites numerous examples of programs and activities, encompassing research, pilot activities, policy guidance, training, and technical assistance. It further found that USAID has collaborated with other organizations in this best practices work, including implementing partners, networks of practitioners, and other donors. USAID appreciates the GAO recognition of the considerable investment it has made in identifying and disseminating best practices, which we hope have helped to advance the field. Concluding Remarks: Given the many barriers confronting poor and very poor households around the world who seek to use MED as a means to improve their economic opportunities and living conditions, USAID has long favored comprehensive support to the microenterprise sector. The major elements of USAID MED programs include efforts to: * expand the availability of financial services to the poor and their enterprises (commonly referred to as "microfinance" and encompassing a range of financial services including but not limited to microcredit); * expand the availability of non-financial enterprise support services such as market access and technical training (commonly referred to as "business development services"); and: improve the policy, legal, regulatory, and administrative environment in which low-income entrepreneurs operate their businesses (commonly referred to as "enabling environment activities"). While support to microcredit is a critical intervention and receives the largest share of the Agency's microenterprise development funding, it is not by itself typically sufficient to help poor and very poor households start and expand enterprises that will move them out of poverty. * The GAO draft report focuses the greatest share of its analysis and observations on microcredit. We would have welcomed a more in-depth review of our entire microenterprise development portfolio, a comparable level of attention to savings, business services and policy activities, and suggestions on strengthening these non-credit programs. We also believe that the effectiveness of USAID's MED efforts should be assessed from the perspective of this broader view of poverty alleviation and reduction. Over the past decade, the development community has come to understand poverty in a more nuanced and comprehensive way, looking beyond a narrow focus on income poverty to understand and intervene on other aspects such as risk, vulnerability, voice and participation. It has also focused attention on the importance of helping the poor build and protect a wide variety of assets (business assets, savings and other financial assets, human capital such as education and health, social capital, etc.). The GAO report focused on a narrow definition: MED programs would meet the test of reducing poverty to the extent that by themselves, they succeeded in moving large numbers of poor and very poor people permanently above the poverty line. This definition applied by the GAO team overlooks or downplays significant impacts that are likely to contribute to poverty reduction, such as reduced risk and vulnerability or empowerment of women and other disadvantaged groups. * MED programs around the world have often contributed measurably to building and protecting assets of poor and very poor among microfinance activities that offered the risk and vulnerability of poor households addition to enterprise credit, this has products (such as emergency loans or school savings, remittances, and micro-insurance. * Some USAID missions have supported MED as a means of strengthening civil society and citizens' voice and stake in the economy. * Other missions have supported MED as a successful intervention to shift from humanitarian to development assistance following conflicts and natural disasters, helping affected populations rebuild assets. While the GAO report mentioned these kinds of activities in passing as evidence of poverty alleviation, it did not indicate their relevance to poverty reduction over time. The draft report concludes that microenterprise development has alleviated poverty but also notes that the evidence is inconclusive on MED's track record on poverty reduction --as measured by large-scale movement of poor people permanently above the poverty line, solely as result of microenterprise assistance. Helping any single household climb out of poverty is a complex process; recent research has demonstrated that progress tends to be uneven and to take a long time. Escaping from poverty usually requires a combination of income asset strategies, and backsliding is possible and perhaps even frequent. USAID does not assert that MED interventions, by themselves, can raise large numbers of very poor people above the poverty line on a permanent basis. * We have always viewed MED as contributing to both poverty alleviation and poverty reduction, by helping poor and very poor households increase their incomes, build up diverse assets, reduce their vulnerability and risk from external shocks, invest in health and education, etc. * Microenterprise assistance can contribute to each of these, by ensuring that services are appropriate, that they reach the intended beneficiaries, and that they are available over time. * Evidence suggests that more powerful poverty impacts come with access to MED services over time; USAID strategies reflect this lesson through a consistent focus on sustainability of service provision, so services can continue beyond the period of donor funding. In sum, while we would agree that empirical evidence (few large-scale longitudinal impact studies with the most rigorous methodologies have been conducted)[NOTE 2] is inconclusive on whether microfinance has reduced poverty, we would not support its summary finding that "there is little evidence that it can lift and keep many over the poverty line" (page 1, emphasis added).her loan fee loans), NOTES: [1] Poverty loans constituted 1.99 million out of a total of 2.9 million loans (p. 32 of the 2001 MRR annual report). [2] One (Khandker, World Bank) analyzed the poverty impact of three large-scale microfinance institutions (MFIs) in Bangladesh. It concluded that on average, 5% of clients per year moved out of poverty as a result of the assistance. Given that these MFls serve millions of clients, one could reasonably argue poverty reduction impact. The following are GAO's comments on USAID's letter dated November 6, 2003. GAO Comments: 1. USAID stated that our report does not address the full range and scope of its microenterprise strategy and program. Our report focuses primarily on microfinance, since this component of USAID's microenterprise program has received, and continues to receive, the bulk of the agency's funding. Microfinance has also been the principal focus of long-term studies funded by USAID and others. We found no long-term studies or evaluations that assessed the impact of USAID's support for Business Development Services (BDS) or its policy work in the area of microenterprise development. Our discussions with USAID employees in Peru, Egypt, and Bulgaria regarding BDS and policy initiatives yielded some information on these efforts, but we found that no data on these efforts had been collected systematically. 2. USAID said that it has long used loan size as a proxy for services to the very poor, recognizing that it is imperfect but a statutory requirement. Because of the limitation of loan size as a proxy, we analyzed impact studies and evaluations funded by USAID and others that collected information on borrowers' economic status to determine the extent to which microfinance has reached the very poor. These studies, based on in-depth research across multiple countries and settings, found that the very poor are rarely reached with micro loans, for reasons outlined in this report (see app. III for a summary of key studies on this topic). To complement information contained in these studies, we discussed this issue at two roundtable meetings with international experts; we also interviewed USAID officials and nongovernmental organization officials working with USAID-funded programs in the countries where we conducted fieldwork. The consensus across the literature and among the experts is that microfinance projects often have difficulty reaching the very poor. 3. USAID said that the MRR has used a single, clear definition of sustainability in questionnaires to implementing partners. We disagree with USAID on this point, and we have added information to this section to clarify our concern regarding the agency's lack of a standardized method for measuring microfinance institutions' (MFI) sustainability. As noted in the report, 38 percent of MFIs that received USAID funding in fiscal year 2001 reported that they had achieved financial sustainability. In addition, the figures cited in USAID's response combine data on operational and financial sustainability, despite the fact that operational sustainability is defined in USAID's policy guidance as an interim measure toward the goal of achieving full financial sustainability. 4. USAID stated that allocating the impact between USAID and other donors would be impractical and methodologically questionable. However, it also says that it plans to include language in its annual report indicating that many of USAID's awardees receive support from other sources as well, and that these sources deserve a share of the credit for the awardees' impacts. 5. See comment 2. USAID states that it made about 2 million loans in fiscal year 2001 that met the statutory standard for service to the very poor. The agency said that it also utilized other methods of reaching this group. As noted in the report, Congress recognized the limitations of loan size as an indicator for targeting and reaching the very poor and directed USAID to develop more accurate methods to ensure that this group is reached in the future. 6. USAID said that the report suggests that loans to the very poor can have negative consequences and may be a significant or widespread problem. As noted in the report, the very poor can benefit from credit, but some evidence suggests that microcredit should compliment, not substitute, for investments in core services, such as health and education. 7. USAID states that their record of supporting MFIs and achieving sustainability is strong. With regard to the issue of MFIs' achieving full financial or operational sustainability, we note in the report that USAID's policy establishes full financial sustainability as its goal; that is, to develop fully financially sustainable MFIs, capable of providing services indefinitely without USAID or other donor support. We did not report data on operational sustainability because this measure is defined in USAID's policy manual as a "useful interim standard of financial performance." Accordingly, we focused on full sustainability, a standard that, if widely attained, could ensure that these institutions would be available to provide these services in the future. Also, see comment 3. 8. USAID said the report suggests that sustainability might not be consistent with serving very poor clients. Our report does not state or suggest that sustainability might not be consistent with serving very poor clients. We agree with USAID that attaining full financial sustainability may be more difficult for MFIs serving greater numbers of very poor borrowers. 9. USAID incorrectly attributed to us an audit conducted in Egypt; this audit was conducted by the USAID Inspector General. 10. USAID stated that its policy allows microenterprise funds to be obligated for activities that do not meet the definition of microenterprise development found in the MRR and that microenterprise awardees do not have to solely serve micro-scale enterprises. However, our report addresses the reporting of such activities, not the policy. According to the 2001 MRR, "Microenterprises are small, often informally organized businesses that are owned and operated by poor and very poor entrepreneurs. USAID defines a microenterprise as one that comprises 10 or fewer employees, including unpaid family workers, in which the owner/operator of the enterprise…is considered poor. By limiting its definition of microenterprises to those whose owners/ operators are poor, USAID ensures that the focus of its efforts remains on the most vulnerable households in higher-risk environments." Despite this definition, the annual MRR reports present data on a wide variety of activities that do not meet this definition. This includes its policy work, much of its BDS work, its obligations to small and medium businesses, and loans to those that are not poor. As a result, it is uncertain how much of USAID's funding is going to poor microenterpreneurs. We believe that USAID should be more transparent in reporting these results. In addition, despite USAID's statement that the MFIs and missions report only activities that meet the definitions of microenterprise as defined in the MRR, we found no evidence of this in our work in three countries or our analysis of MRR data. As noted in this report, we found numerous examples of the missions' and implementing partners' reporting activities to the MRR that did not meet the MRR definition. Based on USAID's comments, we have modified this section of the report to further clarify our position and the basis for these observations. USAID also said it will include more explicit guidance on its website to address the issue. This could potentially improve this aspect of USAID's reporting. 11. USAID stated that the report focused on a narrow definition of program impacts. This report does not take a narrow view of the impacts of USAID's microenterprise program. In addition to our assessment of its impact on poverty alleviation and poverty reduction, there are sections focused on reaching the poor and very poor and other services these groups may need; outreach to women; the sustainability of MFIs; the reliability of the MRR; best practices identified by the microenterprise development industry; USAID's efforts to identify and promote best practices; whether USAID incorporates best practices in their projects; and a synopsis of 22 key studies. In both the body of the report and appendix III, we include considerable discussion of the extent to which microfinance can help alleviate poverty by reducing risk and vulnerability. 12. USAID states that microenterprise development can be a successful intervention to shift from humanitarian to development assistance following conflicts and natural disasters. The USAID policy manual (section II.H.4.), titled "Avoiding Poor Prospects for Microfinance Development," states that microfinance should not be viewed as a response to alleviate the large-scale human suffering created by wars and civil conflict. It notes that such assistance will inevitably conflict with the basic requirements of building sound financial institutions. Despite this guidance, we found that USAID/Bulgaria used emergency funds provided for the Danube River Initiative to respond to the economic hardship resulting from the Kosovo crisis, providing funding to MFIs that committed to work in this region. Officials of the implementing partner told us that this humanitarian initiative, while important from a social perspective, proved to be financially unsustainable in light of the many challenges refugees faced. Accordingly, the implementing partner terminated its programs in these regions, according to USAID officials in Bulgaria. [End of section] Appendix VI: GAO Contact and Staff Acknowledgments: GAO Contact: Phillip Herr (202) 512-8509: Staff Acknowledgments: In addition to the person listed above, Edward George, Jim Strus, Martin De Alteriis, Mona Sehgal, David Dornisch, Yesook Merrill, Valerie Caracelli, and Reid Lowe made key contributions to this report. FOOTNOTES [1] Title I of P.L. 106-309. In legislation enacted on June 17, 2003, Congress amended the act to increase assistance for the poor and provide changes in how assistance is targeted to the very poor (P.L. 108-31). [2] USAID has not defined "poverty reduction." For this report, we applied the definitions found in a 2002 USAID-funded report on microenterprise that defined poverty reduction in two ways: (1) alleviating the impacts of poverty (which can include raising income to permit increased expenditures on necessities such as food, education, and housing); and (2) lifting and keeping a large number of people above a nation's poverty line. [3] The act defines the very poor as "those living in the bottom 50 percent below the poverty line as established by the national government of the country." The act requires that 50 percent of all microenterprise resources be targeted to the very poor. [4] MFIs supply micro loans, savings, and other financial services. To be considered fully sustainable, the MFI must generate sufficient revenues to cover both its operational costs (e.g., salaries) and its financial costs (e.g., the cost of borrowing funds from commercial sources). [5] Micro loans--small loans with frequent repayment schedules and commercial interest rates--are administered by a range of institutions, including nongovernmental organizations and private financial institutions. [6] World Bank, CGAP Report 2000, Consultative Group to Assist the Poorest (Washington, D.C.: 2000). [7] Donald Snodgrass and Jennifer Sebstad, Clients in Context: The Impact of The Microfinance in Three Countries--Synthesis Report (Washington, D.C.: Management Systems International, 2002). [8] Jennifer Sebstad and Monique Cohen, Microfinance, Risk Management, and Poverty (Washington, D.C.: Management Systems International, 2000). [9] USAID generally supports and invests in MFIs and ongoing projects started by NGOs or other groups. The research literature on microfinance and USAID's own studies do not distinguish between USAID- funded projects and projects funded by the World Bank, private foundations, or other donors. [10] Elizabeth Dunn and J. Gordon Arbuckle, Jr., The Impacts of Microcredit: A Case Study from Peru, (Washington, D.C.: USAID/AIMS, 2001). [11] Bashayer El Kheir: An Impact Tracking Study (Final Report) (Washington, D.C.: USAID, 2003). Another USAID study in Egypt was unable to draw firm conclusions about the impact on poverty alleviation, because, although it noted some positive findings, the study lacked baseline data as well as a control group (see USAID, Technical Assistance to Small and Emerging Business Development Organizations: Study on Old Clients [Washington, D.C.: 2001]). [12] Donald Snodgrass and Jennifer Sebstad, Clients in Context: The Impact of The Microfinance in Three Countries--Synthesis Report (Washington, D.C.: Management Systems International, 2002), p. 65. [13] Shahidur Khandker, Fighting Poverty with Microcredit: Experience in Bangladesh. (New York: Oxford University Press, Inc., 1998). [14] Sebstad and Cohen, Microfinance, Risk Management, and Poverty. [15] Sergio Navajas et al., "Microcredit and the Poorest of the Poor: Theory and Evidence from Bolivia," World Development 28, no. 2 (2000): 333-346. [16] PL 108-31, enacted June 17, 2003. [17] Research that attempts to find a middle ground between sustainability-and poverty-orientations is reflected in Morduch (2000), Woller, Dunford, and Woodworth (2000), and Cohen (2002). (See app. III.) [18] Evaluation and Strategic Review of the Consultative Group to Assist the Poor (Washington, D.C.: The World Bank, 2002), p. 16. (See also Rutherford, 2000; Matin, Hulme, and Rutherford, 2002; Cohen, 2002, in app. III.) [19] CGAP, a consortium funded by 29 bilateral and multilateral donor agencies that support microfinance, is affiliated with the World Bank. [20] Donor Brief, No. 13, July 2003, CGAP. [21] USAID applies the poverty lending requirement on an agency wide basis. However, it does not ask that individual missions or MFIs meet this requirement. [22] USAID also reaches a large number of savers through its microfinance program. According to USAID data for 2001, the program reached 3.5 million savers, compared with 2.9 million borrowers. However, the proportion of women reached through micro savings services cannot be determined, because USAID does not collect savings data disaggregated by gender. [23] Several studies indicate that in some cases, micro loans have not empowered women but instead have strengthened traditional gender and class power relations. However, this conclusion does not represent the general agreement among experts we consulted that, on the whole, micro loans have had a positive effect on women. [24] Under the group lending model, 5 to 30 members receive a loan and future loans to individual members are contingent on the entire group's repayment record. [25] USAID provided funding to 492 organizations in fiscal 2001. The agency does not disaggregate these data by MFIs, BDS providers, or policy advocacy organizations. [26] To the extent that financially stable institutions are more likely to report sustainability data, this percentage may be biased upward. [27] In the 2002 survey of 147 MFIs, 42 percent reported full sustainability. The MFIs participated in the survey voluntarily. (See "Bulletin Tables" [data annex], MicroBanking Bulletin 8 (2002). [28] James W. Fox, Mark Havers, and Klaus Maurer, Evaluation and Strategic Review of The Consultative Group to Assist the Poorest (Washington, D.C., The World Bank: 2002), p.iv. [29] Two branches that had been fully sustainable experienced significant declines in sustainability (49 percent and 14 percent, respectively), but managed to remain sustainable, albeit at a lower level. [30] Bulgaria Microfinance Assessment (Arlington, VA: The Peoples' Group, Ltd., 2002), p. 8. [31] USAID, Microenterprise Results Reporting for 2001 (Washington, DC: 2003) p.4. [32] Candace Nelson, Barbara McNelly, Kathleen Stack and Lawrence Yanovitch, Village Banking: The State of the Practice (New York, NY: SEEP Network and the United Nations Development Fund for Women, 1996); Craig Churchill, Madeline Hirschland, and Judith Painter, New Directions in Poverty Finance: Village Banking Revisited (Washington, D.C.: SEEP Network, 2002). [33] Joann Ledgerwood, Microfinance Handbook: An Institutional and Financial Perspective (Washington, D.C.: World Bank, 1998). [34] In group lending and village banking, group members, rather than an individual, accept joint responsibility for repaying the loan. [35] Mohini Malhotra, Poverty Lending and Microenterprise Development: A Clarification of the Issues, GEMINI Working Paper No. 30 (Washington, D.C.: U.S. Agency for International Development, 1992). GAO's Mission: The General Accounting Office, the investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO's commitment to good government is reflected in its core values of accountability, integrity, and reliability. Obtaining Copies of GAO Reports and Testimony: The fastest and easiest way to obtain copies of GAO documents at no cost is through the Internet. GAO's Web site ( www.gao.gov ) contains abstracts and full-text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as "Today's Reports," on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select "Subscribe to e-mail alerts" under the "Order GAO Products" heading. Order by Mail or Phone: The first copy of each printed report is free. Additional copies are $2 each. A check or money order should be made out to the Superintendent of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a single address are discounted 25 percent. Orders should be sent to: U.S. General Accounting Office 441 G Street NW, Room LM Washington, D.C. 20548: To order by Phone: Voice: (202) 512-6000: TDD: (202) 512-2537: Fax: (202) 512-6061: To Report Fraud, Waste, and Abuse in Federal Programs: Contact: Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: fraudnet@gao.gov Automated answering system: (800) 424-5454 or (202) 512-7470: Public Affairs: Jeff Nelligan, managing director, NelliganJ@gao.gov (202) 512-4800 U.S. General Accounting Office, 441 G Street NW, Room 7149 Washington, D.C. 20548: